Cumulative adverse financial circumstances: associations with patient health status and behaviors.

Cumulative adverse financial circumstances: associations with patient health status and behaviors. There are approximately 39.4 visits per 100 people to U.S. hospitalemergency departments (EDs) annually (Niska, Bhuiya, & Xu, 2010).Across health service disciplines, there is a sizeable, long-standingbody of evidence establishing the role of the ED as a health systemsafety net for vulnerable populations, emphasizing the need for socialwork response to patients' psychosocial and economic needs in thismedical setting (Bergman, 1976; Gordon, 1999, 2009; Gordon, Chudnofsky,& Hayward, 2001; Healy, 1981; Walls, Rhodes, & Kennedy, 2002).In general, the current responsibilities of social workers in U.S. EDsinclude addressing physician-identified social service needs, counselingand crisis intervention, discharge planning, and referral to relevantservices (Auerbach & Mason, 2010; Holliman, Dziegielewski, &Datta, 2001). Typically, these direct, individualized interactions arebrief, single-session contacts (Kitchen & Brook, 2005). Although theneed for patient-level social work functions in EDs is well-researched,the value of system-level social health screening and response in EDshas yet to be established (Auerbach & Mason, 2010; Gibbons &Plath, 2005; Gordon, 2001; Keehn, Roglitz, & Bowden, 1994; McCoy,Kipp, & Ahern, 1992; Ponto & Berg, 1992). The result is thatroutine screening and referral for patient economic deprivation (forexample, food insecurity, housing instability, unemployment or incomesecurity, lack of adequate health coverage for medications and physiciancare) are scarce practices in contemporary medical settings (Fleegler,Lieu, Wise, & Muret-Wagstaff, 2007). The relationship between socioeconomic well-being and biomedicaland behavioral health is well-documented (Adler et al., 1994; Bosma,Schrijvers, & Mackenbach, 1999; Head & Faul, 2008; Krieger,Williams, & Moss, 1997; Lynch, Kaplan, Cohen, Tuomilehto, &Salonen, 1996; Oakes & Rossi, 2003; Schrijvers, Stronks, van deMheen, & Mackenbach, 1999). In the United States, a patient'ssocioeconomic status (SES) is a multilayered construct that includesindividual-, household-, and neighborhood-level consideration of socialclass, race and ethnicity, gender, education level, and income andaccess to financial resources (Krieger et al., 1997). Although no singleSES factor explains the connection between SES and health, one componentthat is malleable is an individual's experience of financialconstraints (Fleegler et al., 2007). Similarly, among the many othernon-SES factors associated with disease and disability that lie outsidethe reach of public policy, individual financial constraints can beimmediately addressed through broad health policies, physician treatmentstrategies (that is, prescriptions for less costly medications), andtargeted social work referrals to existing public programs (Kaplan &Lynch, 2001; Piette, Heisler, & Wagner, 2004; Poleshuck & Green,2008; Schrijvers et al., 1999). Currently, the U.S. health system is entering a critical juncturein its development. The present fiscal crisis has resulted in 46 of 50U.S. states facing budget shortfalls for fiscal year 2011, and thenumber of unemployed Americans has reached 14.9 million, with 8.9million Americans who want to work full-time but are underemployedworking part-time (McNichols & Johnson, 2010; U.S. Bureau of LaborStatistics, 2010). In the current political climate, the economicrecession was accompanied by considerable federal support to statehealth agencies, unprecedented funding to health services researchers,and comprehensive legislative reforms to the health system. Federalstimulus money granted through the American Recovery and ReinvestmentAct of 2009 (P.L. 111-5) includes $750 million in contracts to buildfully integrated electronic medical record (EMIL) infrastructures inhealth systems nationwide; $225 million in funding for healthinformation technology (IT) job training programs; and, beginning inOctober 2011, $14 to $27 billion in bonuses to physicians and hospitalsif they demonstrate "meaningful use" of their EMIL systems(Ferris, 2010; Sack, 2009; U.S. Department of Health and Human Services,Office of the Secretary, 2009). In addition, the Patient Protection andAffordable Care Act (ACA) (P.L. 111-148) was signed into law on March23, 2010. Among the many reforms emphasized in this landmark health careoverhaul legislation are training and research support for improvingpatient-centered communication and care coordination and paymentincentives for centralized care delivery through voluntary accountablecare organizations that focus on coordination and quality of care(Levinson, Lesser, & Epstein, 2010; Shortell, Casalino, &Fisher, 2010). The present system transition toward a full-scale healthinformatics infrastructure (in which the definition of "meaningfuluse" will be developed over the next several years) marks a uniqueopportunity to incorporate social work screening and response as anessential feature of routine standard of care in acute health caresafety-net settings. In this context, understanding the current burdenof ED patients' adverse financial circumstances and theirassociation with patient health status and health behaviors can informwhether there is a pressing need for fully integrated, system-levelsocial work screening and intervention. OBJECTIVES Accordingly, the present investigation was designed to examine theprevalence of patient-disclosed food insecurity, housing instability,employment concerns, and lack of adequate health coverage formedications and physician care. In addition, we sought to determinewhether cumulative adverse financial circumstances are relevant to EDpatient health status and behaviors. METHOD Participants This is a secondary analysis of a prospective, cross-sectionalstudy of a convenience sample of nonemergent adult patients presentingto the ED at an urban, tertiary care teaching hospital between May andOctober 2009. This ED had a volume of approximately 61,962 patientvisits in 2009. Patients who presented to the ED with any nonemergentcomplaint that did not require an urgent intervention were 18 years orolder and were wilting and able to complete a survey voluntarilyself-administered an anonymous 40-item paper-pencil Social Health Survey(SHS) that was previously validated through cognitive interviewing withseveral hundred nonemergent, urban ED patients from diverse backgroundsto ensure content and construct validity (Rhodes et al., 2001). Clerksin the ED were instructed to distribute surveys 24 hours a day, 7 days aweek to all patients on arrival in the ED if the patient was notimmediately directed to an urgent medical intervention. Survey questionswere written in English, and only patients who were able to read English(fifth-grade reading level) were eligible to participate. The patientsreturned the survey to the provider or nurse caring for them in the ED.After review by the ED provider, the anonymous surveys were placed in acollection box and patients' responses were entered into a databasefor further analysis. The study protocol was reviewed and approved bythe institutional review board (IRB), which considered this anonymouspatient data exempt from IRB oversight. Measures Demographics/Control Variables. Age (18 to 25, 26 to 35, 36 to 45,46 to 65, and over 65), gender (male versus female), race (white, black,Hispanic, Asian, multiracial, other), and education level (less thanhigh school, high school, some college, completed college or more) werecollected using the SHS. The surveys were self-administered, and someanswers were left blank. On these control variables, multiple imputationmethods were used to account for this missingness in logistic regressionanalyses. Exposure to Economic Deprivation. Five dichotomous (disclosedversus not disclosed) variables of adverse financial circumstances werestudied as patient exposure to economic deprivation: food insecurity,housing instability, employment concerns, cost-related medicationnonadherence (CRMN), and cost barriers to accessing physician care.Participants were asked, "Do you have concerns about any of thefollowing?" followed by a list of 19 check-box options thatincluded the following: not enough food, housing, employment, andcan't afford medication. In addition, participants were asked thefollowing three questions: 1. Do you sometimes eat less than you would like because of moneyconcerns? 2. Have you ever not seen a doctor when you needed to because ofmoney concerns? 3. Have you ever not taken medications you needed because of moneyconcerns? To meet criteria for experiencing food insecurity, participants hadto disclose either having concerns about not enough food or sometimeseating less because of money concerns. Participants who reported havingconcerns about housing or employment were considered positive forexposure to housing instability and employment concerns. Participantswho reported either having a concern about not being able to affordmedication or ever not taking medications due to money concerns wereconsidered positive for exposure to CRMN. If participants reported evernot seeing a doctor due to money concerns, this was considered apositive disclosure of experiencing cost barriers to physician care.Responses of not sure to the three specific questions of foodinsecurity, cost barriers to physician care, and CRMN were consideredpositive disclosures. Missing information on any of these items wasconsidered a nondisclosure. After assessment of whether participants hadscreened positive for each individual category, the cumulative number ofadverse financial circumstances was calculated (range: 0 to 5). Patient Health Status/Health Behaviors. An item that questioned howparticipants would rate their own overall health (with response optionsof poor, fair, good, very good, or excellent) was used to measureoverall patient health status. Prior research has established that thisquestion of self-rated health is a strong predictor of mortality(DeSalvo, Bloser, Reynolds, He, & Muntner, 2006). Consistent withprior literature, self-rated health responses were dichotomized intocategories of poor/fair versus good/very good/excellent (Felitti et al.,1998). As depression is the leading contributor to the global burden ofdisease, an item gauging level of depressed mood was measured (WorldHealth Organization, 2010). Any answer of yes or not sure to thisquestion, "In the last 12 months have you felt sad or depressedmore than two weeks in a row?" was considered a positive disclosureof depressed mood. Although the effects of psychological stress onhealth have not been fully explicated, there is evidence that highstress has an effect on physiological processes that leads to manycommon diseases (Ismail, Winkley, & Rabe-Hesketh, 2004; Pickering,2001; Segerstrom & Miller, 2004; Yuen, Thompson, Flugel, Bell, &Sander, 2007). An item questioned participants about how much stressthey were under, with response options of none, just a little, normalamount, too much, or extreme. Responses were dichotomized into none/justa little/normal amount versus too much/extreme. Three behavioral riskfactors that are known contributors to the leading causes of morbidityand mortality in the United States served as outcome variables: smokingtobacco, excessive alcohol consumption, and illicit drug use (Mokdad,Marks, Stroup, & Gerberding, 2004). For smoking tobacco and usingillicit drugs, participants were asked, "Have you smoked ANYcigarettes in the last 12 months?" and "Have you used ANYstreet drugs in the last four weeks?" Responses of yes or not surewere considered positive disclosures of tobacco or illicit drug use. Tomeasure unsafe alcohol consumption, participants were asked the numberof days per week that they have an alcoholic drink and the number ofdrinks they have on a typical drinking day. National Institute onAlcohol Abuse and Alcoholism (NIAAA) gender-specific criteria forat-risk drinking (that is, exceeding seven drinks per week for women, 14drinks per week for men) were used to create a dichotomous variable ofwhether the participant screened positive for unsafe alcohol consumption(NIAAA, 2010). In addition, heavy episodic drinking behavior (four ormore drinks in one sitting) was measured with a question askingparticipants how many times in the last year they have had four or moredrinks in one day (NIAAA, 2010). For all dependent variables, multipleimputation methods were used to account for any missing data in logisticregression analyses. Statistical Analyses All data entry and statistical analyses were performed using SPSSversion 17.0 (SPSS Inc., Chicago). Data were analyzed descriptively tocharacterize the demographic characteristics (age, gender, race, andeducation level) of the overall sample and the subgroups of participantswith cumulative adverse financial circumstances. Logistic regressionanalysis was used to adjust for the potential confounding effects ofdemographic variables on the relationship between patients' numberof adverse financial circumstances and health problems. A"dose-response" relationship of adverse financialcircumstances to self-rated health and behavioral health risks wastested by entering the number of adverse financial circumstances as asingle ordinal variable (0, 1, 2, 3, 4, and 5) into a separate logisticregression model for health status and behavioral risk factor dependentvariable. To account for missing data on control and dependentvariables, multiple imputation methods were used to produce correctstandard errors and consistent estimates of regression coefficients(Allison, 2010). Adjusted odds ratios and 95 percent confidenceintervals were calculated. RESULTS Demographics The demographic characteristics of the total sample and forsubgroups of cumulative exposures to adverse financial circumstances aredescribed in Table 1. The 1,506 patients who completed the SHS werepredominantly female (65 percent), and the mean age of the total samplewas 37 years (range: 18 to 90). Compared with the demographiccharacteristics of this urban ED's general patient population (N =61,962) for that year (January to December 2009), the group ofnonemergent patients who participated in this study was distinct in thatit was less elderly (over 65 years of age) (5 percent versus 11percent), less black (39 percent versus 65 percent), less white (16percent versus 29 percent), and more female (65 percent versus 58percent). In this sample, 384 (25.5 percent) patients disclosed CRMN,352 (23.4 percent) disclosed experiencing barriers to physician care,346 (23.0 percent) disclosed food insecurity, 290 (19.3 percent)disclosed employment concerns, and 272 (18.1 percent) disclosed housinginstability. Slightly more than half of respondents (51.9 percent, n =781) were unexposed to adverse financial circumstances; 257 (17.1percent) had one exposure, 182 (12.1 percent) had two exposures, 157(10.4 percent) had three exposures, 93 (6.2 percent) had four exposures,and 36 (2.4 percent) had five exposures. Taken together, 725 (48percent) of the 1,506 respondents reported one or more exposures toadverse financial circumstances, and 468 (31 percent) reported two ormore exposures to adverse financial circumstances. Relationship between Exposures to Economic Deprivation and Health Logistic regression models (which included age, gender, race, andeducational attainment as covariates) found that as patients'number of exposures to adverse financial circumstances increased, theirrisk (adjusted odds ratio) increased for poor/fair self-rated health,depressed mood, high stress, smoking, and illicit drug use (see Table2). For these outcome variables, there was a significant dose-responserelationship between the number of adverse financial circumstances andhealth risks. When participants with five categories of exposure werecompared with those with none, the adjusted odds ratios ranged from 24.7for high stress to 3.4 for poor/ fair self-rated health (see Table 2).An exception to this pattern was that the risk (adjusted odds ratio) ofunsafe alcohol consumption and heavy episodic drinking did notdemonstrate the same graded relationship and did not significantlyincrease as the number of exposures to adverse financial circumstanceincreased. DISCUSSION Almost half of our sample of urban ED patients reported at leastone exposure to economic deprivation, and almost a third reported two ormore exposures to economic deprivation. The two most common adversefinancial circumstances among these ED patients were an inability toafford needed medications due to cost and an inability to see a doctorwhen needed due to cost. We found a striking dose response between thenumber of exposures to economic deprivation and multiple markers of riskfor the leading causes of disease and death in adults. Compared withpatients unexposed to economic deprivation, patients who had experiencedfive categories of adverse financial circumstances had a threefoldincrease in poor/fair self-rated health, a 17-fold increase in depressedmood, a 24-fold increase in high stress, a sixfold increase in smoking,and a fivefold increase in illicit drug use. These findings suggest thatthe impact of economic deprivation on patient health status is strongand cumulative. Contrary to prior findings linking financial stress to increasedalcohol use (Dawson, Grant, & Ruan, 2005; San Jose, van Oers, van deMheen, Garretsen, & Mackenbach, 2000), patients with cumulativeadverse financial circumstances were not more or less likely to reportunsafe alcohol consumption. It is possible that unmeasured factors--suchas family alcohol abuse, childhood exposure to abuse and householddysfunction, marital status, and depressive symptoms--are more stronglyassociated with unhealthy drinking behavior (Felitti et al., 1998;Merrick et al., 2008). These possible explanations for the absence of arelationship between patient-reported financial problems and hazardousalcohol use behavior warrant further study. These results must be interpreted with consideration of severalimportant limitations. First, our study setting was an urban adultteaching hospital ED, and, therefore, the results might not begeneralizable to children or other settings. Notably, only 2.7 percentof our sample's respondents reported having an education levelbelow high school, which is lower than the 8.0 percent dropout rate for16- to 24-year-olds in the general U.S. population (U.S. Department ofEducation, National Center for Education Statistics, 2010). Even withinour setting, our convenience sample of ED patients differs from the basepopulation of ED patients, which makes it difficult to generalizeresults from the SHS to the setting's general patient population.Second, convenience sampling is vulnerable to selection bias if patientswho did not choose to complete the survey or patients who were noteligible to complete the survey (for example, those who were visuallyimpaired, illiterate, or nonproficient English readers) were more, orless, likely to be positive for adverse financial circumstances ormarkers of health risks. Third, some people with poor health status andnegative health behaviors may be more, or less, likely to report adversefinancial circumstances, which limits inferences of causality.Similarly, there may be important mediators or moderators of therelationship between adverse financial circumstances and health statusthat were unmeasured in our study. Nonetheless, there is reason tobelieve that our estimates are conservative and the relationship betweenthese variables may actually be stronger due to patients'reluctance to disclose sensitive financial concerns and stigmatizedadverse health behaviors. Moreover, the large number of patients willingto self-disclose these risks represents "low-hanging fruit"deserving of social work intervention. IMPLICATIONS FOR RESEARCH AND POLICY The enactment of health care reform marks an opportunity to remoldU.S. health care service delivery processes (Sebelius, 2010) and toexpand the paradigm of health services to more directly account forsocial determinants of health. As the implementation stages of the ACAcommence, policymakers are charged with the tasks of developing nationalquality improvement strategies and establishing guidelines for the useof health IT (White House, President Barack Obama, 2010). In thisprocess, it will be possible to organize and advocate for arestructuring of the extent to which social work practices and servicesare integrated into the U.S. health infrastructure. Critical prioritiesfor these restructuring efforts include upgrading the qualifications anddefinitions of social workers in health systems, expanding the hours ofsocial work coverage and the clinical (patient-level) component ofsocial work practice in medical settings, and promoting the inclusion ofsocial work departments in the overall health system organization (NASW,2009). In addition to these important policy goals, the current studysignals the need for extensive broad-based, system-level integration ofsocial work services. Our findings reinforce the call for routinepatient self-assessment of economic constraints in all medical settings(Fleegler et al., 2007) and specifically underscore the need for thesepractices in the ED setting. It is reasonable to urge social work professionals to weigh in onthe implementation of health care reform. A longitudinal survey ofsocial work administrators in hospital settings found a reportedincrease in the "social work influence" in hospital strategicdecision making and system reorganization (Mizrahi & Berger, 2005).Empirical evidence can better equip social work leaders in theircommunication with hospital decision makers and those draftingregulations to accompany legislative reforms (Alexander, Hearld, Jiang,& Fraser, 2007). Therefore, further research using rigorousmethodologies are needed to examine the effect of hospital-based systeminterventions on reducing ED patient economic deprivation and theresulting effect on clinical care quality. Given the current allocationof federal resources toward the conversion toward EMRs, interventionsthat leverage ITs for social health screening are of particular researchinterest. Outcome measures of societal costs and costs to providers andhealth systems are also necessary to secure funding for patient carethat includes consideration of patient financial distress. Original manuscript received September 28; 2010 Accepted November 12, 2010 REFERENCES Adler, N. E., Boyce, T., Chesney, M.A., Cohen, S., Folkman, S.,Kahn, R. L., & Syme, S. L. (1994). Socioeconomic status and health:The challenge of the gradient. American Psychologist, 49, 15-24. Alexander, J. A., Hearld, L. R., Jiang, H.J., & Fraser, I.(2007). Increasing the relevance of research to health care managers:Hospital CEO imperatives for improving quality and lowering costs.Health Care Management Review, 32, 150-159. Allison, P. D. (2010). Missing data. In P. Marsden & J. Wright(Eds.), Handbook of survey research (Rev. 2nd ed., pp. 631-658).Bingley, England: Emerald Group Publishing Limited. American Recovery and Reinvestment Act of 2009, P.L. 111-5, 123Stat. 115. Auerbach, C., & Mason, S. E. (2010). The value of the presenceof social work in emergency departments. Social Work in Health Care, 49,314-326. Bergman, A. S. (1976). Emergency room: A role for social workers.Health & Social Work, 1(1), 32-44. Bosma, H., Schrijvers, C., & Mackenhach, J. P. (1999).Socioeconomic inequalities in mortality and importance of perceivedcontrol: Cohort study. BMJ (Clinical Research Ed.), 319, 1469-1470. Dawson, D. A., Grant, B. F., & Ruan, W.J. (2005). Theassociation between stress and drinking: Modifying effects of gender andvulnerability Alcohol & Alcoholism, 40, 453-460. DeSalvo, K. B., Bloser, N., Reynolds, K., He, J., & Muntner, P.(2006). Mortality prediction with a single general self-rated healthquestion. A meta-analysis. Journal of General Internal Medicine, 21,267-275. Felitti, V.J., Anda, R. F., Nordenberg, D., Williamson, D. F.,Spitz, A. M., Edwards, V., et al. (1998). Relationship of childhoodabuse and household dysfunction to many of the leading causes of deathin adults: The Adverse Childhood Experiences (ACE) study. AmericanJournal of Preventive Medicine, 14, 245-258. Ferris, N. (2010). Health IT gold rush under way. Health Affairs,29, 583-584. Fleegler, E.W., Lieu, T. A.,Wise, P. H., & Muret-Wagstaff, S.(2007). Families' health-related social problems and missedreferral opportunities. Pediatrics, 119, e1332-e1341. Gibbons, J., & Plath, D. (2005). "Everybody puts a lotinto it!" Single session contacts in hospital social work. SocialWork in Health Care, 42(1), 17-34. Gordon, J. A. (1999). The hospital emergency department as a socialwelfare institution. Annals of Emergency Medicine, 33, 321-325. Gordon, J. A. (2001). Cost-benefit analysis of social work servicesin the emergency department: A conceptual model. Academic EmergencyMedicine: Official Journal of the Society for Academic EmergencyMedicine, 8, 54-60. Gordon, J.A. (2009). Social welfare in the hospital emergencydepartment: The challenge of integrated community health care. Annals ofEmergency Medicine, 53, 603-604. Gordon, J.A., Chudnofsky, C. R., & Hayward, R.A. (2001). Wherehealth and welfare meet: Social deprivation among patients in theemergency department. Journal of Urban Health: Bulletin of the New YorkAcademy of Medicine, 78(1), 104-111. Head, B.A., & Faul, A. C. (2008). Development and validation ofa scale to measure socioeconomic well-being in persons with cancer.Journal of Supportive Ontology, 6, 183-192. Healy, J. (1981). Emergency rooms and psychosocial services. Health& Social Work, 6(1), 36-43. Holliman, D. C., Dziegielewski, S. F., & Datta, P. (2001).Discharge planning and social work practice. Social Work in Health Care,32(3), 1-19. Ismail, K., Winkley, K., & Rabe-Hesketh, S. (2004). Systematicreview and meta-analysis of randomised controlled trials ofpsychological interventions to improve glycaemic control in patientswith type 2 diabetes. Lancet, 363, 1589-1597. Kaplan, G., & Lynch, J. (2001). Is economic policy healthpolicy? American Journal of Public Health, 91, 351-353. Keehn, D. S., Roglitz, C., & Bowden, M. L. (1994). Impact ofsocial work on recidivism and non-medical complaints in the emergencydepartment. Social Work in Health Care, 20(1), 65-75. Kitchen, A., & Brook, J. (2005). Social work at the heart ofthe medical team. Social Work in Health Care, 40(4), 1-18. Krieger, N., Williams, D. R., & Moss, N. E. (1997). Measuringsocial class in US public health research: Concepts, methodologies, andguidelines. Annual Review of Public Health, 18, 341-378. Levinson, W., Lesser, C. S., & Epstein, R. M. (2010).Developing physician communication skills for patient-centered care.Health Affairs, 29, 1310-1318 Lynch, J. W., Kaplan, G.A., Cohen, R. D., Tuomilehto, J., &Salonen, J.T. (1996). Do cardiovascular risk factors explain therelation between socioeconomic status. risk of all-cause mortality,cardiovascular mortality, and acute myocardial infarction? AmericanJournal of Epidemiology, 144, 934-942. McCoy, H.V., Kipp, C.W., & Ahern, M. (1992). Reducing olderpatients' reliance on the emergency department. Social Work inHealth Care, 17(1), 23-37. McNichols, E., & Johnson, N. (2010). States continue to feelrecession's impact. Retrieved from the Center on Budget and PolicyPriorities Web site: http://www.cbpp.org/cms/index.cfm?fa=view&id=711 Merrick, E. L., Horgan, C. M., Hodgkin, D., Garnick, D.W.,Houghton, S. F., Panas, L., et al. (2008). Unhealthy drinking patternsin older adults: Prevalence and associated characteristics. Journal ofthe American Geriatrics Society, 56, 214-223. Mizrahi, T., & Berger, C. S. (2005). A longitudinal look atsocial work leadership in hospitals: The impact of a changing healthcare system. Health & Social Work, 30, 155-165. Mokdad, A. H., Marks, J. S., Stroup, D. F., & Gerberding, J. L.(2004). Actual causes of death in the United States, 2000. JAMA, 291,1238-1245. doi:10.1001/ jama.291.10.1238 National Association of Social Workers. (2009). Health care policy.In Social work speaks: National Association of Social Workers policystatements, 2009-2012 (pp. 167-170). Washington, DC: Author. National Institute on Alcohol Abuse and Alcoholism. (2010)Rethinking drinking: Alcohol and your health. Retrieved fromhttp://rethinkingdrinking.niaaa.nih.gov/ Niska, R.W., Bhuiya, F., & Xu, J. (2010). National HospitalAmbulatory Medical Survey: 2007 Emergency department summary, nationalhealth statistics reports No. 26, August 6, 2010. Hyattsville, MD:National Center for Health Statistics. Oakes, J. M., & Rossi, P. H. (2003). The measurement of SES inhealth research: Current practice and steps toward a new approach.Social Science & Medicine, 56, 769-784. Patient Protection and Affordable Care Act, P.L. 111-148, 124 Stat.119-1025 (2010). Pickering, T. G. (2001). Mental stress as a causal factor in thedevelopment of hypertension and cardiovascular disease. CurrentHypertension Reports, 3, 249-254. Piette, J. D., Heisler, M., & Wagner, T. H. (2004).Cost-related medication underuse among chronically ill adults: Thetreatments people forgo, how often, and who is at risk. American journalof Public Health, 94, 1782-1787. Poleshuck, E. L., & Green, C. R. (2008). Socioeconomicdisadvantage and pain. Pain, 136, 235-238. Ponto, J. M., & Berg, W. (1992). Social work services in theemergency department: A cost-benefit analysis of an extended coverageprogram. Health & Social Work, 17, 66-73. Rhodes, K.V., Lauderdale, D. S., Stocking, C. B., Howes, D. S.,Roizen, M. F., & Levinson, W. (2001). Better health while you wait:A controlled trial of a computer-based intervention for screening andhealth promotion in the emergency department. Annals of EmergencyMedicine, 37, 284-291. Sack, K. (2009, July 19). Defying slump, 13 states insure morechildren. New York Times, A1. San Jose, B., van Oers, H.A.M., van de Mheen, H. D., Garretsen,H.F.L., & Mackenbach, J. P. (2000). Stressors and alcoholconsumption. Alcohol and Alcoholism, 35, 307-312. Schrijvers, C.T., Stronks, K., van de Mheen, H. D., &Mackenhach, J. P. (1999). Explaining educational differences inmortality: The role of behavioral and material factors. American Journalof Public Health, 89, 535-540. Sebelius, K. (2010, July 26, 2010). Ensuring success: Must focus onpatients, eliminating inefficiencies. Retrieved fromwww.commonwealthfund.org Segerstrom, S. C, & Miller, G. E. (2004). Psychological stressand the human immune system: A meta-analytic study of 30 years ofinquiry. Psychological Bulletin, 130, 601-630. Shortell, S. M., Casalino, L. P., & Fisher, E. S. (2010). Howthe center for Medicare and Medicaid innovation should test accountablecare organizations. Health Affairs, 29, 1293-1298. U.S. Bureau of Labor Statistics. (2010). Employment situationsummary--August 2010. Retrieved from http://www.bls.gov/news.release/empsit.nr0.htm U.S. Department of Education, National Center for EducationStatistics. (2010). The condition of education 2010 (NCES 2010-028),Table A-19-2. Retrieved fromhttp://nces.ed.gov/fastfacts/display.asp?id=16 U.S. Department of Health and Human Services, Office of theSecretary. (2009). Department of Health and Human Services: Notice ofavailability of federal matching shares for Medicaid and foster care andadoption assistance, April 21, 2009. Federal Register, 74, 18236-18237. Walls, C.A., Rhodes, K.V., & Kennedy, J.J. (2002). Theemergency department as usual source of medical care: Estimates from the1998 national health interview survey. Academic Emergency Medicine:Official Journal of the Society for Academic Emergency Medicine, 9,1140-1145. White House, President Barack Obama. (2010). The Affordable CareAct--Implementation timeline. Retrieved fromhttp://www.whitehouse.gov/healthreform/ timeline World Health Organization. (2010). Depression. Retrieved fromhttp://www.who.int/mental_health/ management/depression/definition/en/ Yuen, A. W.C., Thompson, P.J., Flugel, D., Bell, G. S., &Sander, J.W. (2007). Mortality and morbidity rates are increased inpeople with epilepsy: Is stress part of the equation? Epilepsy &Behavior, 10(1), 1-7. Joanna Bisgaier, MSW, is a doctoral student, School of SocialPolicy & Practice, University of Pennsylvania, Philadelphia. KarinV. Rhodes, MD, MS, is director, Division of Emergency Care & PolicyResearch, Department of Emergency Medicine, University of PennsylvaniaSchool of Medicine and School of Social Policy & Practice, 3815Walnut Street, Philadelphia, PA I9104; e-mail: kvr@sp2.upenn.edu. Partsof an earlier version of this article appeared on a poster at theSociety for Social Work and Research meeting, January 15, 2011, Tampa,FL.Table 1: Prevalence of Cumulative Adverse FinancialCircumstances, by Demographic Characteristics Percent Exposed to Number of Categories (a) Total Sample 0 1 2Characteristic (N = 1,506) (n = 781) (n = 257) (n = 182)Age group (years) 18-25 25.8 26.8 25.7 25.3 26-35 24.2 22.4 24.5 33.0 36-45 13.9 12.0 14.0 10.4 46-65 19.3 19.7 19.5 19.2 > 65 5.1 7.0 5.4 3.3 Missing data 11.8 12.0 10.9 8.8Gender Male 27.6 27.4 28.0 24.7 Female 64.6 64.7 63.8 70.3 Missing data 7.8 7.9 8.2 4.9Race White 15.5 15.5 13.2 10.4 Black 39.2 39.2 44.4 46.2 Hispanic 1.4 1.4 1.9 1.6 Asian 2-5 2.5 2.3 1.6 Multiracial 2.3 2.3 1.6 3.8 Other 1.2 1.2 0.4 2.2 Missing data 37.9 37.9 36.2 34.1Education < High school 2.7 2.4 4.7 2.2 High school 21.8 16.8 24.9 30.8 Some college 18.2 16.8 19.1 18.7 > College grad 14.7 20.1 11.7 9.3 Missing data 42.6 43.9 39.7 39.0 Percent Exposed to Number of Categories (a) 3 4 5Characteristic (n = 157) (n = 93) (n = 36)Age group (years) 18-25 21.0 23.7 33.3 26-35 19.7 24.7 33.3 36-45 21.0 24.7 11.1 46-65 23.6 12.9 8.3 > 65 0.6 1.1 0.0 Missing data 14.0 12.9 13.9Gender Male 27.4 31.2 33.3 Female 62.4 59.1 63.9 Missing data 10.2 9.7 2.8Race White 7.6 15.5 15.5 Black 40.1 39.2 39.2 Hispanic 0.0 1.4 1.4 Asian 0.0 2.5 2.5 Multiracial 6.4 2.3 2.3 Other 1.9 1.2 1.2 Missing data 43.9 37.9 37.9Education < High school 1.9 3.2 0.0 High school 26.1 30.1 22.2 Some college 21.0 19.4 25.0 > College grad 5.1 5.4 11.1 Missing data 45.9 41.9 41.7* 0 = respondents with zero adverse financial circumstancesreported; 1 = respondents with one category of adverse financialcircumstances; 2 = respondents with two categories of adversefinancial circumstances reported; 3 = respondents with threecategories reported; 4 = respondents with four categoriesreported; 5 = respondents with all five categories of adversefinancial circumstances reported.Table 2: Number of Categories of Adverse Financial Circumstancesand the Odds of Health Problems, Adjusted for Age, Gender, Race,and Educational Attainment (N = 1,506) AdjustedHealth Problems and Number of Prevalence OddsHealth Risk Categories (%) (a) Ratios (b)BehaviorsPoor/fair self-rated 0 41.2 1.0health 1 16.4 1.2 2 14.1 1.7 * 3 15.3 2.5 * 4 9.5 3.1 * 5 3.5 3.4 *Depressed mood (in 0 23.5 1.0 *last two weeks, most 1 23.5 3.7 *of the time) 2 14.2 2.7 * 3 19.3 5.8 * 4 13.0 7.1 * 5 6.5 17.9 *High stress (under 0 31.0 1.0"too much" or 1 20.8 2.8 *extreme stress) 2 15.4 2.9 * 3 15.2 3.9 * 4 11.2 6.4 * 5 6.4 24.7 *Tobacco use 0 39.2 1.0(in last 12 months) 1 19.5 1.6 * 2 13.1 1.5 * 3 13.8 2.3 * 4 9.7 2.9 * 5 4.7 5.9 *Unsafe alcohol 0 40.6 1.0consumption (exceeds 1 28.1 2.1 *gender-specificNIAAA safe alcohol 2 6.3 0.6consumption 3 10.9 1.3criteria) (c] 4 12.5 2.4 * 5 1.6 0.8Heavy episodic 0 46.1 1.0drinking (d) (in 1 21.3 1.7 *past 12 months) 2 13.1 1.3 3 10.1 1.5 4 7-1 1.4 5 2.2 0.9Illicit drug use 0 33.3 1.0(in last four weeks) 1 20.2 1.8 * 2 14.0 1.7 3 14.9 2.1 * 4 10.5 2.5 * 5 7.0 5.0 * 95%Health Problems and ConfidenceHealth Risk IntervalBehaviorsPoor/fair self-rated Referenthealth 0.9-1.7 1.2-2.5 1.7-3.7 1.9-4.9 1.7-7.1Depressed mood (in Referentlast two weeks, most 2.6-5.4of the time) 1.8^.0 3.8-8.7 4.4-11.7 8.1-39.3High stress (under Referent"too much" or 2.0-3.9extreme stress) 2.1-4.2 2.6-5.7 4.0-10.2 9.3-65.2Tobacco use Referent(in last 12 months) 1.2-2.2 1.1-2.2 1.6-3.3 1.8-4.5 2.7-12.5Unsafe alcohol Referentconsumption (exceeds 1.1-4.1gender-specificNIAAA safe alcohol 0.2-1.8consumption 0.5-3.1criteria) (c) 1.0-6.0 0.1-6.2Heavy episodic Referentdrinking (d) (in 1.1-2.6past 12 months) 0.8-2.2 0.9-2.6 0.8-2.5 0.3-2.3Illicit drug use Referent(in last four weeks) 1.0-3.0 0.9-3.1 1.1-4.0 1.2-5.0 2.0-12.2Note: NIAAA = National Institute on Alcohol Abuse and Alcoholism.(a) Prevalence percentages based on nonmissing data.(b) Odds ratios, p values, and confidence intervals derived usingmultiple imputation for missing data.(c) More than seven drinks per week for women and more than 14drinks per week for men.(d) Four drinks or more on a single occasion.* p < .05.