Volume 9, Issue 1 , Pages 58-63, January 2012
The Impact of Socioeconomic Status and Comorbid Medical Conditions on Ionizing Radiation Exposure From Diagnostic Medical Imaging in Adults
Article Outline
Purpose
The aim of this study was to characterize cumulative exposure to ionizing radiation from diagnostic imaging (CEDI) in adult patients and investigate its relationship to a patient's socioeconomic status and comorbid medical conditions.
Methods
A retrospective cohort study was conducted of 54,015 patients seen within the outpatient clinic system of an academic, tertiary care, urban medical center during the month of January 2006, estimating the CEDI from all procedures performed within 3 years of the index visit (until January 2009). Socioeconomic status was estimated from census tract geocoding. Comorbid medical conditions were identified from the electronic medical record.
Results
A total of 9,537 adult patients were seen within the index month and underwent imaging tests. Eighty percent were living in areas with >10% poverty. Thirty-six percent of men and 43% of women had diagnoses from the Elixhauser list. Mean CEDI values were 10 ± 19.09 mSv for those from areas with >10% poverty and 8.9 ± 22.42 mSv for those living in areas with <10% poverty. Poverty and comorbidities covaried. Estimated CEDI within groups of patients with the same comorbidity was not associated with socioeconomic status.
Conclusion
At this institution, there is a high prevalence of patients living in poverty. Those living in poverty are at higher risk for comorbid conditions that are associated with increased CEDI. However, controlling for comorbidity, socioeconomic status was no longer predictive of CEDI.
Key Words: Radiation dose , radiation exposure , exposure to patients and personnel , socioeconomic factor , access to health care
Introduction
The use of both ionizing radiation and nonionizing imaging technology to diagnose and treat disease has transformed medical practice over the past half century. Although the benefits of imaging are accepted by most, the side effects of ionizing radiation exposure from CT scans, fluoroscopy, and nuclear medicine studies are receiving more attention. A recent study indicated that 2% of nonelderly adults in the United States are exposed to ionizing radiation from medical imaging procedures at levels >20 mSv/y, the dose restriction recommended by industry because of concerns of solid tumor and leukemia development [1]. Medical imaging accounts for 48% of the US population's ionizing radiation exposure, and CT contributes one-half of this medical imaging radiation dose [2]. The use of CT has increased rapidly, with an estimated 70 million scans performed in 2007 in the United States [3]. On the basis of epidemiologic data from atomic bomb survivors, it is estimated that 1.5% to 2% of future cancers in the United States may be attributable to current CT use and that 29,000 cancers may be attributable to CT studies performed in 2007 alone [4, 5].
The amount of cumulative exposure to ionizing radiation from diagnostic imaging (CEDI) varies widely according to specific patient characteristics such as age and gender [1]. Lower socioeconomic status (SES), lack of health insurance, and belonging to a disadvantaged race or ethnic group are associated with increased disease prevalence, decreased access to care, and worse health outcomes across a broad spectrum of diseases [6, 7, 8, 9, 10].
Understanding the factors contributing to the utilization of radiologic imaging is important in correcting health disparities regarding increased radiation exposure and limited access to nonionizing diagnostic imaging alternatives such as MRI and ultrasound. Associations between diagnostic imaging utilization, SES, and medical radiation exposure in the US health care setting have not been reported. A recent study estimating radiation exposure from medical imaging in the United States was limited to patients with private insurance [1]. Our study population, primarily African American and Latino, represents adult patients aged <40 years seen at an urban medical center in the United States. Our hypothesis postulates an association between estimates of CEDI and SES, controlling for patient comorbidities and other demographic characteristics.
Methods
Design
This study was approved by the medical center's institutional review board for the protection of human subjects and was compliant with HIPAA. A retrospective cohort was derived from patients who accessed care at a tertiary academic urban medical center with outpatient, inpatient, and emergency facilities. We accessed our institution's computerized medical record system using Clinical Looking Glass version 3.3 (Montefiore Medical Center, Bronx, New York), an interactive software application, to capture radiation exposure, geocoding, comorbidities, demographics, and insurance.
Study Population
The index population included all patients aged 21 to 40 years at the time of an initial visit during January 2006 at any of the institution's outpatient sites that use the electronic medical record system. The cutoff at 40 years of age was chosen because the adverse effects of radiation dramatically decrease as one approaches this age [4]. Each record was reviewed through January 2009 to identify all medical imaging studies performed over this time period. Patients who died during the 3 years of follow up were excluded to minimize potential bias due to numerous examinations preceding their deaths and truncation of their observation period [1].
Demographic Information
Age, gender, race, and ethnicity were self-reported by patients at registration. Insurance information was based on the source of payment recorded for the original outpatient encounter and was subsequently categorized as private insurance, Medicare, Medicaid, or no insurance.
SES
The percentage of people living below the poverty level in a census tract has been used as a measure of SES [11, 12]. In an attempt to validate these data derived from this cohort, 100 randomly selected addresses geocoded by the Clinical Looking Glass geocoding report were compared with the census tract on the US Census Bureau's geocoding Web site [13]. We found that 82% of addresses were assigned the same census tract by both methods; 10% could not be geocoded by the Census Bureau's Web site, and a small fraction, 8%, were assigned a different census tract. To account for the possible nonlinearity of the relationship between census tract percentage poverty and radiation exposure, the categories of 0% to 10%, >10% to 20%, >20% to 30%, >30% to 40%, >40% to 50%, and >50% were created. Bronx County has one the highest poverty rates in the nation, 28.3%, and the study population therefore did not replicate the cutoff of >20%, which is the federal definition of a poverty area, as the highest poverty group [12].
Examination Utilization and Estimation of Radiation Dose
All diagnostic radiology examinations, nuclear medicine examinations, and cardiac catheterizations were recorded for 3 years from the index outpatient visit date for each patient. These included all procedures performed at multiple imaging facilities, including inpatient, emergency, and outpatient settings. A mean radiation dose was assigned to the common examinations performed in radiology, nuclear medicine, and invasive cardiology on the basis of literature-reported values (mammography, CT, MRI, ultrasound, nuclear medicine, and cardiac catheterization) [14, 15]. For each patient, the estimated radiation doses for all examinations for the 3-year period were then summed, yielding a total estimated cumulative radiation dose (estimated CEDI). Actual measured radiation exposures vary widely and also tend to be higher than our estimated mean calculated exposures [3, 16].
Comorbidities
Given that SES is associated with disease risk, comorbidities were incorporated into analyses to account for increases in imaging procedures due to increased disease burden. The presence of each of the 30 Elixhauser diagnoses, a comprehensive set of comorbidity measures for use with large administrative inpatient data sets, for an individual patient at any time over the 3-year study period was determined by searching International Classification of Diseases, ninth rev, codes. The Elixhauser diagnoses have been shown to be positively associated with mortality and hospital charges [17].
Statistical Analysis
For men and women separately, descriptive statistics are presented as mean ± SD or as medians and ranges as appropriate for continuous variables and as relative frequencies for categorical variables. Bivariate analyses between percentage poverty and diagnosis and between diagnosis and estimated CEDI were performed using Wilcoxon's rank-sum tests. Associations between percentage poverty and estimated CEDI were performed using Spearman's rank correlations. Multiple linear regression analysis with a monitored backward variable elimination procedure was used to derive models of the relationship between cumulative radiation exposure and patient characteristics, with estimated cumulative radiation dose as the dependent variable. Because the distribution of estimated cumulative radiation was not normal, transformations of scale were attempted to better approximate assumptions of normally distributed error terms. However, with most patients receiving very low doses or no radiation, and the data set sufficiently large, errors were reasonably normal, and thus assumptions of the multiple linear regression analyses were not violated. Bivariate analyses between CEDI and demographics and insurance were performed using Kruskal-Wallis tests for categorical or short-scale ordinal variables or Wilcoxon's rank-sum tests for dichotomous variables. Variables significantly associated with CEDI were included in multivariate models. The analysis was done using Medicare and Medicaid data as well as insurance data (results not shown), and because insurance was highly associated with poverty and results were similar, reported results were based on percentage poverty instead of insurance status. Primary analyses included CEDI as the dependent variable and age and its effect on poverty, and poverty and its effect on diagnosis, as independent variables. Sensitivity analyses were performed for the entire data set with CEDI as the dependent variable and the following independent variables: age, gender, ethnicity, race, census tract percentage poverty as either continuous or categorical variables, insurance categories, and diagnosis, as well as within insurance group. Variables retained in final models were those significant at P < .05. Analyses were performed using SAS version 9.1.2 (SAS Institute Inc, Cary, North Carolina).
Results
Characteristics of the study population are detailed in Table 1. Of the 54,015 patients in the cohort, there were 9,537 adults (2,323 men, 7,214 women; age range, 22-40 years; mean age, 31 ± 5 years) who underwent imaging studies within the index month. Approximately 50% were Hispanic, 36% African American, 10% white, and 2% Asian. Nearly two-thirds were living in areas with >20% poverty, and almost 80% were living in areas with >10% poverty. The degree of poverty in which patients lived was associated with receiving Medicaid, with 8.21% of those on Medicaid living in areas of <10% poverty compared with 21.80% of those not on Medicaid (P < .0001).
Table 1. Distribution of demographic characteristics, insurance, and diagnosis for adult men and women
| Variable | Men Aged 22-40 y (n = 2,323) | Women Aged 22-40 y (n = 7,214) |
|---|---|---|
| Age (y) | 31.35 ±5.34 | 30.61±5.25 |
| Race | ||
| American Indian | 7(0.30%) | 38(0.53%) |
| Asian | 47(2.02%) | 122(1.69%) |
| African American | 748(32.20%) | 2,678(37.12%) |
| Multiracial | 237(10.20%) | 1,079(14.96%) |
| Pacific Islander | 3(0.13%) | 8(0.11%) |
| White | 257(11.06%) | 664(9.20%) |
| Ethnicity | ||
| Hispanic | 730(51.19%) | 2,820(53.69%) |
| Insurance | ||
| Medicaid | 669(28.80%) | 2,916(40.42%) |
| Medicare | 99(4.26%) | 143(1.98%) |
| None | 288(12.40%) | 703(9.74%) |
| Private | 1,267(54.54%) | 3,452(47.85%) |
| Percentage poverty | ||
| <10% | 450(19.62%) | 1,163(16.30%) |
| 10% to <20% | 415(18.10%) | 1,162(16.29%) |
| 20% to <30% | 451(19.67%) | 1,469(20.59%) |
| 30% to <40% | 397(17.31%) | 1,345(18.86%) |
| 40% to <50% | 430(18.75%) | 1,430(20.05%) |
| ≥50% | 150(6.54%) | 564(7.91%) |
| Diagnoses | ||
| Myocardial infarction | 8(0.39%) | 14(0.21%) |
| Congestive Heart Failure | 25(1.23%) | 51(0.76%) |
| Peripheral vascular disorders | 13(0.64%) | 32(0.47%) |
| Cerebrovascular disease | 22(1.08%) | 60(0.89%) |
| Dementia | 0(0.00%) | 1(0.01%) |
| Chronic pulmonary disease | 254(12.51%) | 1,320(19.58%) |
| Peptic ulcer disease | 16(0.79%) | 28(0.42%) |
| Mild liver disease | 72(3.55%) | 114(1.69%) |
| Diabetes without complications | 181(8.91%) | 371(5.50%) |
| Hemiplegia or paraplegia | 28(1.38%) | 22(0.33%) |
| Moderate or severe liver disease | 83(4.09%) | 257(3.81%) |
| Metastatic solid tumor | 7(0.34%) | 13(0.19%) |
| Valvular disease | 15(0.74%) | 52(0.77%) |
| Pulmonary circulation disorders | 10(0.49%) | 44(0.65%) |
| Complicated hypertension | 76(3.74%) | 129(1.91%) |
| Hypothyroidism | 28(1.38%) | 219(3.25%) |
| Lymphoma | 10(0.49%) | 19(0.28%) |
| RA CVD | 12(0.59%) | 113(1.68%) |
| Coagulopathy | 26(1.28%) | 93(1.38%) |
| Fluid and electrolyte disorders | 78(3.84%) | 182(2.70%) |
| Blood loss anemia | 4(0.20%) | 11(0.16%) |
| Deficiency anemia | 84(4.14%) | 878(13.02%) |
| Drug abuse | 83(4.09%) | 133(1.97%) |
| Psychoses | 146(7.19%) | 431(6.39%) |
| Depression | 177(8.71%) | 858(12.72%) |
Among men, 30.44% of those who lived in areas with <10% poverty underwent at least one ionizing radiation imaging study, compared with 35.76% living in areas of greater poverty (P = .0338); the overall means for estimated CEDI were 2.71 ± 13.01 and 3.58 ± 12.38 mSv, respectively (P = .0261). Among those with at least one ionizing radiation imaging study, the means were 8.91 ± 22.42 and 10.01 ± 19.09 mSv, respectively (P = .4710). In men receiving Medicaid, 43.50% received ionizing radiation, compared with 31.26% for those not on Medicaid (P < .0001); the difference in overall means was 5.60 ± 16.42 vs 2.57 ± 10.45 mSv, respectively (P < .0001). Men on Medicaid or living in areas of >10% poverty were more likely to have congestive heart failure (P = .015), cerebrovascular disease (P = .0027), chronic pulmonary disease (P < .0001), hemiplegia or paraplegia (P < .0001), complicated hypertension (P = .0108), lymphoma (P = .0056), rheumatoid arthritis (P = .0053), coagulopathy (P = .0019), fluid and electrolyte disorders (P < .0001), deficiency anemia (P < .0001), drug abuse (P < .0001), psychosis (p<.0001), and depression (P < .0001).
Importantly, within all Elixhauser subgroups, CEDI varied independently of SES measures. Diagnoses in men associated with percentage poverty or Medicaid and increased estimated CEDI included congestive heart failure, cerebrovascular disease, chronic pulmonary disease, hemiplegia or paraplegia, complicated hypertension, lymphoma, coagulopathy, fluid and electrolyte disorders, deficiency anemia, drug abuse, psychosis, and depression. The prevalence among men with one or more of these diagnoses associated with estimated CEDI was 27.59%. In bivariate analyses, mean estimated CEDI differed significantly by Medicaid status for chronic pulmonary disease (P = .0101), deficiency anemia (P = .0162), psychosis (P = .0451), and depression (P = .0009), with those on Medicaid having higher mean estimated CEDI; no further associations were identified for percentage poverty. In multivariate analyses controlling for age, significant associations between either poverty or Medicaid and estimated CEDI remained for chronic pulmonary disease and depression only.
Among women, 36.80% of those who lived in areas with <10% poverty underwent at least one ionizing radiation imaging test, compared with 40.47% living in areas of greater poverty (P = .0194); the difference in overall means was 2.90 ± 11.21 vs 4.11 ± 15.62 mSv (P = .0053). Among those with at least one ionizing radiation test, the difference in means was 7.87 ± 17.40 vs 10.15 ± 23.27 mSv (P = .0466). In those on Medicaid, 39.23% received ionizing radiation, compared with 40.20% among those not on Medicaid (P = .4074); the difference in overall means was 4.75 ± 16.59 versus 3.30 ± 13.64 mSv, respectively (P = .6694).
Within all Elixhauser subgroups, CEDI varied independently of SES measures. Women on Medicaid or living in areas of >10% poverty were more likely to have cerebrovascular disease (P = .0153), chronic pulmonary disease (P < .0001), uncomplicated diabetes (P = .0247), complicated hypertension (P = .0012), pulmonary circulation disorders (P = .0362), rheumatoid arthritis (P = .0002), coagulopathy (P = .0002), fluid and electrolyte disorders (P < .0001), blood loss anemia (P = .0063), deficiency anemia (P < .0001), drug abuse (P < .0001), psychosis (P < .0001), and depression (P < .0001). In women, the following diagnoses were associated with Medicaid or percentage poverty and increased estimated CEDI: cerebrovascular disease, chronic pulmonary disease, diabetes without complications, pulmonary circulation disorders, complicated hypertension, rheumatoid arthritis, coagulopathy, fluid and electrolyte disorders, blood loss anemia, deficiency anemia, drug abuse, psychosis, and depression. In women, the prevalence of one or more of these diagnoses was 39.95%. In bivariate analyses, mean estimated CEDI differed significantly by Medicaid status for uncomplicated diabetes (P = .0038), complicated hypertension (P = .0163), fluid and electrolyte disorders (P < .0001), and deficiency anemia (P = .0313), with larger mean estimated CEDI for those on Medicaid. In addition, percentage poverty was significantly correlated with estimated CEDI for those with cerebrovascular disease (P = .0242). All diagnoses except for cerebrovascular disease were associated with increased estimated CEDI after controlling for age in multivariate analyses.
Discussion
In this study, we found that living in greater poverty is associated with increased CEDI. However, our results attempt to explain this relationship by controlling for increased comorbidities among patients of lower SES. Importantly, within each Elixhauser diagnosis group controlling for age and sex, SES was no longer associated with CEDI.
A total of 36.07% of men and 43.28% of women had at least one of the diagnoses on the Elixhauser list. Contrary to expectations pertaining to limited access to care among those of lower SES, this study demonstrates that those patients living in census tracts with greater poverty or who were Medicaid recipients had higher CEDI. The sample represents an urban multiracial and multicultural population, with 30% of subjects living at or below the poverty level. Previously described gender-related differences in prevalence of disease were also reproduced in the sample.
Subjects living in greater poverty or who were on Medicaid had more disease than those living in lower poverty areas and, in turn, those with more disease had higher CEDI. However, there were relatively few Elixhauser diagnoses for which the association between poverty or Medicaid status and estimated CEDI was significant after controlling for age; these subsets include men with chronic pulmonary disease or depression and women with uncomplicated diabetes, complicated hypertension, fluid and electrolyte disorders, or deficiency anemia.
For men, differences in estimated CEDI were more attributable to whether they were receiving Medicaid rather than the degree of poverty of the area in which they lived. Men who were on Medicaid were significantly more likely to accumulate higher CEDI values than those not on Medicaid, and there was no association between CEDI exposure and percentage poverty of their home environment. For women, the reverse was true; there was no difference in estimated CEDI by Medicaid status, but there was with regard to percentage poverty of their home environment.
Two previous studies investigated the relationship between SES and diminished access to medical imaging in large populations. A Canadian study found that patients in the highest income quintile were more likely to undergo nearly all radiologic examinations [18]. However, a Taiwanese study found that lower SES was associated with a higher rate of CT utilization [19]. Neither of these studies reported cumulative radiation exposure estimates.
Although the assertion that a significant number of cancers are caused by medical radiation has been questioned [20, 21], the need to minimize unnecessary ionizing radiation has been widely accepted. The “American College of Radiology White Paper on Radiation Dose in Medicine” cites research indicating a significant cancer increase at radiation levels >50 mSv and notes that it would not be uncommon for patients receiving multiple CT scans to have an estimated exposure above this level [16]. Because of these concerns, the International Commission on Radiological Protection recommends that occupational effective radiation doses be limited to an effective dose of 100 mSv over 5 years, with a maximum of 50 mSv in any year [22].
Study Limitations
Limitations of this study include possible inaccuracies in census tract geocoding and the limited ability of population-based census tract poverty data and Medicaid status to approximate the SES of individual patients. Similarly, the high prevalence of impoverished subjects within the study population required the stratification of large groups; the majority of these could be classified as poor. Those within the lowest SES category and at greatest risk may be underrepresented, as they may fail to obtain Medicaid because of access issues. Those in extreme poverty and unable to access the hospital clinic system are similarly underreported. An additional limitation is that examinations were performed within a single hospital system's multiple imaging facilities, and examinations performed at other institutions were not captured. Thus, our results likely underestimate cumulative radiation exposure. Furthermore, our findings may be limited to factors unique to our institution or its patient population. The study cohort had a greater proportion of African Americans, had a smaller proportion of Caucasians, and was poorer in relation to the overall Bronx population. In addition, our analyses within select Elixhauser diagnostic groups may not have accounted for all relevant comorbidities, which could bias associations between poverty and estimated CEDI in these groups. Finally, estimates and not actual doses of CEDI were obtained. As noted, actual doses vary widely on the basis of patient and scanner characteristics. The use of shielding is not taken into account, and repeat scans for technically poor scans were not counted.
Conclusions
Medical imaging provides valuable information in appropriate settings, and many tests can only be done using ionizing radiation. Most authors agree that exposure to ionizing radiation, even at levels from diagnostic testing, is associated with an increased risk for solid tumor formation. This retrospective cohort study supported results from previous work showing that patients of lower SES have greater disease burden, which likely leads to more imaging, causing further accumulation of CEDI. This study identified some Elixhauser diagnoses that were highly prevalent in the population sample (chronic obstructive pulmonary disease, diabetes, depression, psychosis). Targeted interventions, such as educating treating physicians about nonionizing alternative tests or encouraging radiologists to suggest these alternatives at the time tests are ordered, may reduce health disparities. For all other diagnoses, strategies targeted to those living in impoverished areas or to those receiving Medicaid are required to prevent not only disease but exposure to ionizing radiation and its longer term risk for cancer. The retrospective nature of this study did not allow us to address variables affecting utilization.
References
- Exposure to low-dose ionizing radiation from medical imaging procedures . N Engl J Med . 2009;361:849–857
- . Scientific Committee 6-2 on Radiation Exposure of the U.S. Population (Ionizing radiation exposure of the population of the United States (Report No 160)) . Bethesda, Md: National Council on Radiation Protection and Measurements; 2009;
- Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer . Arch Intern Med . 2009;169:2078–2086
- . Computed tomography—an increasing source of radiation exposure . N Engl J Med . 2007;357:2277–2284
- Projected cancer risks from computed tomographic scans performed in the United States in 2007 . Arch Intern Med . 2009;169:2071–2077
- . Inequality in quality: addressing socioeconomic, racial, and ethnic disparities in health care . JAMA . 2000;283:2579–2584
- . Unequal treatment: confronting racial and ethnic disparities in health care . Washington, DC: National Academy Press; 2003;
- . Health insurance coverage and the risk of decline in overall health and death among the near elderly 1992-2002 . Med Care . 2006;44:277–282
- . Explaining US racial/ethnic disparities in health declines and mortality in late middle age: the roles of socioeconomic status, health behaviors, and health insurance . Soc Sci Med . 2006;62:909–922
- Socioeconomic differences in the prevalence of common chronic diseases: an overview of eight European countries . Int J Epidemiol . 2005;34:316–326
- . Geocoding and monitoring of US socioeconomic inequalities in mortality and cancer incidence: does the choice of area-based measure and geographic level matter? (The Public Health Disparities Geocoding Project) . Am J Epidemiol . 2002;156:471–482
- . Painting a truer picture of US socioeconomic and racial/ethnic health inequalities: the Public Health Disparities Geocoding Project . Am J Public Health . 2005;95:312–323
- . American Fact Finder: advanced geography search . http://factfinder.census.gov/servlet/AGSGeoAddressServlet?_lang=en&_programYear=50&_treeId=420 Accessed November 15, 2009
- . Effective doses in radiology and diagnostic nuclear medicine: a catalog . Radiology . 2008;248:254–263
- Radiation exposure from medical imaging in patients with chronic and recurrent conditions . J Am Coll Radiol . 2010;7:351–359
- American College of Radiology white paper on radiation dose in medicine . J Am Coll Radiol . 2007;4:272–284
- . Comorbidity measures for use with administrative data . Med Care . 1998;36:8–27
- . Socioeconomic status and the utilization of diagnostic imaging in an urban setting . CMAJ . 2005;173:1173–1177
- . Disease patterns and socioeconomic status associated with utilization of computed tomography in Taiwan 1997-2003 . J Formos Med Assoc . 2008;107:145–155
- . Cancer risk from low-level radiation . AJR Am J Roentgenol . 2002;179:1137–1143
- . The linear no-threshold relationship is inconsistent with radiation biologic and experimental data . Radiology . 2009;251:13–22
- . New ICRP recommendations . J Radiol Protect . 2008;161:
This article is available for CME credit. Visit www.acr.org, ACR Education, online learning for more information.
PII: S1546-1440(11)00392-9
doi:10.1016/j.jacr.2011.07.009
© 2012 American College of Radiology. Published by Elsevier Inc. All rights reserved.
Volume 9, Issue 1 , Pages 58-63, January 2012