If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Radiology, University of Washington, Seattle, WADepartment of Harborview Injury Prevention & Research Center, University of Washington, Seattle, WA
Department of Medicine, University of Washington, Seattle, WADepartment of Center for Cost and Outcomes Research, University of Washington, Seattle, WA
Department of Radiology, University of Washington, Seattle, WADepartment of Health Services, University of Washington, Seattle, WADepartment of Neurological Surgery, University of Washington, Seattle, WADepartment of Center for Cost and Outcomes Research, University of Washington, Seattle, WA
Relyea-Chew A, Hollingworth W, Chan L, Comstock BA, Overstreet KA, Jarvik JG. Personal bankruptcy after traumatic brain or spinal cord injury: the role of medical debt.
Objective
To estimate the prevalence of medical debt among traumatic brain injury (TBI) and spinal cord injury (SCI) patients who discharged their debts through bankruptcy.
Design
A cross-sectional comparison of bankruptcy filings of injured versus randomly selected bankruptcy petitioners.
Setting
Patients hospitalized with SCI or TBI (1996–2002) and personal bankruptcy petitioners (2001–2004) in western Washington State.
Participants
Subjects (N=186) who filed for bankruptcy, comprised of 93 patients with previous SCI or TBI and 93 randomly selected bankruptcy petitioners.
Interventions
Not applicable.
Main Outcome Measures
Medical and nonmedical debt, assets, income, expenses, and employment recorded in the bankruptcy petition.
Results
Five percent of randomly selected petitioners and 26% of petitioners with TBI or SCI had substantial medical debt (debt that accounted for more than 20% of all unsecured debts). SCI and TBI petitioners had fewer assets and were more likely to be receiving government income assistance at the time of bankruptcy than controls. SCI and TBI patients with a higher blood alcohol content at injury were more likely to have substantial medical debts (odds ratio=2.70; 95% confidence interval, 1.04–7.00).
Conclusions
Medical debt plays an important role in some bankruptcies after TBI or SCI. We discuss policy options for reducing financial distress after serious injury.
TRAUMATIC BRAIN INJURY and SCIs often lead to high out-of-pocket costs and loss of income for patients and their families. These costs may deter patients from seeking rehabilitative care and, in the extreme, lead to bankruptcy.
Immediately after the tightening of bankruptcy laws in 2005, the number of personal bankruptcies decreased by more than 50%. However, the most recent data suggest that an increase in bankruptcies resumed in 2007.
Credit deregulation, job volatility, home-ownership costs, medical care costs, family breakup, and less social stigma associated with bankruptcy are potential factors contributing to the rise of bankruptcy.
The prevalence of medically related bankruptcy is particularly controversial and has important policy implications. Previous research has produced mixed findings. Fay et al
of the bankruptcy filings of more than 6000 patients with TBI and SCI, we found a 33% increase in bankruptcy incidence during the first 5 years postinjury. In other work,
surveys of bankruptcy filers indicate that a substantial proportion (28%) attributes their financial plight to ill health or injury. However, individuals may overemphasize the role of ill health because it is a relatively socially acceptable reason for debt accumulation.
Before drawing conclusions from these studies, it is important to disentangle the potential pathways that lead from ill health to bankruptcy. Figure 1 provides a conceptual framework; pathways include direct medical costs (eg, medical bills not covered by insurance), indirect costs (eg, lost income not covered by disability or social security insurance), and other nonmedical costs (eg third-party injury or property-damage claims).
Fig 1Potential pathways between injury and personal bankruptcy. †Employer-sponsored health insurance.
The primary objective of this study was to estimate the prevalence of bankruptcy with substantial medical debt comparing bankruptcy petitioners with TBI and SCI with a cohort of randomly selected bankruptcy petitioners. We aimed to provide information on the debate on the relative importance of direct medical costs, indirect costs, and nonmedical costs in the pathway between injury and bankruptcy.
Methods
SCI and TBI patients were selected from the Seattle, WA, Harborview Medical Center trauma registry as described previously.
In brief, subjects were eligible if they were hospitalized between June 1, 1996, and December 31, 2002, with a primary or ancillary International Classification of Diseases–9th Revision–Clinical Modification diagnosis code indicating new nonpenetrating TBI or SCI; they were 20 years or older when injured; they resided in western Washington State; and they had a valid SSN recorded in the trauma registry. Trauma registry data were then linked to the Western District of Washington bankruptcy court database by using SSN and patient first and last name. We identified a cohort of injured patients who filed for bankruptcy within 5 years after hospitalization between June 1, 2001, and December 31, 2004. We excluded incomplete bankruptcy petitions. We excluded bankruptcies filed after 2004 because the bankruptcy law changed in late 2005. The choice of a 5-year posthospitalization cutoff is based on our previous work
showing that bankruptcy incidence peaks in the second year postinjury and returns toward preinjury levels by 5 years postinjury. We included bankruptcies filed under either U.S. Code Chapter 7 (Liquidation) or Chapter 13 (Adjustment of Debts of an Individual With Regular Income). In Chapter 7 cases, any nonexempt assets are sold and proceeds are distributed among creditors. In Chapter 13 cases, debtors commit to repayment of all secured and priority debts and to a full or partial repayment of unsecured debts.
For comparison, we randomly selected an equal number of petitioners from a list of all individuals filing for bankruptcy in the same district during the same time period. We crosschecked the randomly selected petitioners to confirm that none of them had been hospitalized with TBI or SCI at our institution during the study period.
For patients with SCI and TBI, the trauma registry included data on demographics, ISS, GCS,
blood alcohol level, length of stay, hospital charges, functional independence at discharge from acute care, acute care discharge disposition (eg, home, SNF, rehabilitation unit), and health care insurer. The ISS is an anatomic scoring system for measuring the severity of the initial trauma, ranging from 1 (lowest) to 75 (unsurvivable).
Factors associated with higher levels of injury severity in occupants of motor vehicles that were severely damaged in traffic crashes in Kentucky, 2000-2001.
we categorized ISS as mild/moderate (<15) or severe/critical (≥15). Functional independence was recorded at the time of discharge by using the modified FIM.
The modified FIM measures independence on a 4-point scale for 3 dimensions (feeding, locomotion, and expression). Overall scores range from 3 (full assistance required with feeding, locomotion, and expression) to 12 (fully independent).
We abstracted data from the bankruptcy petition on the value of real estate and other personal assets, secured debts (eg, a mortgage), unsecured priority debts (eg, taxes or alimony), unsecured nonpriority debts (eg, credit card), and unsecured debts owed to medical or dental service providers. We also recorded the number of dependents, employment duration, current income, current total monthly expenses, and monthly medical and dental expenses reported in the bankruptcy petition.
Creditors' names and a brief description of the source of the debt are recorded in the bankruptcy petition for each unsecured debt. Because medical service providers often use debt-collection agencies, the creditors' names can be ambiguous. Therefore, we classified medical debt as any item in which the creditor name or the described source of debt indicated hospital, ambulatory, dental, nursing facility, or home health care services. Two authors (ARC, WH) unaware of injury status abstracted data from the bankruptcy petitions.
We calculated the ratio of medical debts to all unsecured debts. In a bankruptcy petition in which 100% of unsecured debt is medical debt, it is likely that the bankruptcy would not have occurred in the absence of medical bills. However, there is no consensus on a threshold proportion above which a bankruptcy could be deemed medical debt related. In the health expenditure literature, the definition of “catastrophic health expenditure” has varied between 5% and 20% of total household income.
Thus, as a starting point, we defined “bankruptcy with substantial medical debt” as any bankruptcy petition in which medical debts exceed 20% of all unsecured debts. In sensitivity analyses, we varied this threshold from 10% to 50%.
We derived a measure of earned income stability based on the primary debtor's response to questions about employment duration and income. We coded people as having stable income if they met either of the following 2 conditions: (1) they had worked for the same employer for more than 12 months and earned more than $500 per month net of payroll taxes, excluding social security and other government assistance, and (2) they did not report duration of employment but earned more than $500 net per month, excluding social security and other government assistance. This multilevel definition is necessary because some disabled individuals report long employment duration even though, currently, they are unable to attend work and do not receive salary. In addition, many self-employed individuals do not report employment duration but do have high current and previous year income.
Distributions of assets, income, expenses, and debts are skewed by a few extremely high values; therefore, we report median values and IQRs. We compared the distribution of assets, income, expenses, and debts of the randomly selected petitioners to petitioners with injuries by using the Mann-Whitney U test. Means and proportions were compared with the Student t test and Pearson chi-square test, respectively. We stratified analyses of petitioners with injuries by discharge disposition (home vs SNF or rehabilitation unit) to examine whether the location of postacute care was associated with wealth at the time of bankruptcy. We tested whether TBI and SCI patients were more likely to have bankruptcy with substantial medical debt than the randomly selected petitioners. In subgroup analyses among only those with SCI and TBI, we used forward stepwise binary logistic regression to explore whether the prevalence of bankruptcy with substantial medical debt was associated with demographic or injury-related characteristics. Specifically, we included age group (<30y, 30–40y, >40y), sex, race (white or nonwhite), presence of alcohol (not measured/below legal limit [.08], above), discharge disposition (home vs SNF or rehabilitation unit), TBI, ISS (<15 vs ≥15), FIM score (<12 vs 12 or not recorded), and medical care sponsor (Medicaid, commercial, other) as independent variables. Variables were entered into the model if P was less than .10. All analyses were performed by using SPSS version 14.0
SPSS, 233 S. Wacker Dr, 11th Fl, Chicago, IL 60606.
for Windows. The University of Washington Institutional Review Board approved all study methods.
Results
Of 4372 patients in the trauma registry who met our clinical inclusion criteria, 93 (2.1%) filed for bankruptcy between June 2001 and December 2004. Table 1 presents the demographic and clinical characteristics of these 93 patients stratified by discharge disposition after acute inpatient treatment. Injured petitioners had a mean age of approximately 36, and the majority was white and male. Most had either commercial or Medicaid medical care insurance. Injured petitioners who were discharged to a rehabilitation unit or an SNF had worse GCS, ISS, and FIM scores; were hospitalized for approximately 11 days longer; spent 3.3 days more in the intensive care unit; and had acute care hospital charges totaling $50,000 more than injured petitioners who were discharged to home (all P<.01). Thirty-nine percent of petitioners who were discharged to a rehabilitation unit or an SNF had SCI compared with 2% of those who were discharged to home (P<0.01).
Table 1Demographic and Clinical Characteristics of Injured Patients Who Filed for Bankruptcy Stratified by Discharge Disposition
There were relatively few significant differences between the bankruptcy petitions of injured and randomly selected petitioners (table 2). Compared with randomly selected petitioners, injured petitioners reported fewer total assets ($11,150 vs $24,550, P<.01) and fewer secured debts ($4,600 vs $12,050; P=.03) at the time of bankruptcy. They were also more likely to report receiving assistance from social security or other government programs (33% vs 17%, P=.01). Compared with injured petitioners discharged to their own home, injured petitioners discharged to an SNF or rehabilitation unit were more likely to report their marital status as divorced or separated at the time of bankruptcy (33% vs 13%, P=.02) and to receive social security or other government assistance (49% vs 25%, P=.02). All groups had similar levels of monthly household income and expenses reported on the bankruptcy petition. There was also no evidence that injured petitioners had higher ongoing monthly medical and dental expenses at the time of the bankruptcy.
Table 2Bankruptcy Filings by Injury Status and Discharge Disposition
The majority of randomly selected petitioners reports no medical debts in the bankruptcy petition (median, $0; range, $0–$9010). Median medical debts were higher in injured petitioners who were discharged home after acute care (median, $1044; range, $0–$120,000) and highest in injured petitioners who were discharged to an SNF or rehabilitation unit (median, $1773; range, $0–$68,200). Twenty seven percent of injured petitioners discharged to an SNF or rehabilitation unit, 25% of injured petitioners discharged home, and 5% of randomly selected petitioners had bankruptcy with substantial medical debt based on our primary threshold (table 3). However, this finding is sensitive to the threshold used; for example, only 5% of those with TBI and SCI and 1% of randomly selected petitioners had medical debts that made up more than 50% of their unsecured debts.
Table 3Medical Debts by Injury Status and Discharge Disposition
Proportion of Unsecured Debt That is Medical Debt
Randomly Selected Petitioners (n=93) (%)
Injured Petitioners Discharged Home (n=60) (%)
Injured Petitioners Discharged to SNF/Rehabilitation Unit (n=33) (%)
Stepwise logistic regression showed little association between patient demographics, injury characteristics, and the odds of having substantial medical debts (≥20% of unsecured debts) at bankruptcy. Age, sex, race, ISS, FIM score, discharge disposition, TBI, and medical insurer were not statistically significantly associated with high medical debt. There was some evidence that patients who were over the legal alcohol limit for driving were more likely to have substantial medical debts at bankruptcy (odds ratio=2.70; 95% confidence interval, 1.04–7.00) than patients who were not tested or were under the limit. The proportion of those with TBI and SCI with high levels of medical debt was similar between groups with Medicaid or commercial insurance at the time of injury (fig 2).
Fig 2Medical debt as a proportion of all unsecured debt in patients with Medicaid (n=29) or commercial insurance (n=48) at the time of injury.
We found that many individuals who are hospitalized with TBI or SCI file for bankruptcy with substantial levels of medical debt soon after the injury. Of the 93 TBI and SCI patients in this study, 24 (26%) had medical debt that amounted to more than 20% of all unsecured debt at the time of bankruptcy. This suggests that medical debt is an important contributor to bankruptcy for some patients. It is possible that more comprehensive medical care insurance would have prevented bankruptcy in some of these 24 cases.
Direct Medical Debts
There has been relatively little research examining the amount of medical debt in bankruptcy petitions. Himmelstein et al
found that 27% of a sample of households filing for bankruptcy reported that they had at least $1000 in uncovered medical debt in the previous 2 years. Applying an identical threshold to the petitioners in this study, we also found that 27% of the randomly selected petitioners (and 54% of those with TBI and SCI) had more than $1000 in medical debt. We believe that a $1000 threshold would overstate the prevalence of medical debt–related bankruptcy after brain injury or SCI. Small amounts of medical debt may not be a cause of financial insolvency but rather a symptom of wider financial problems.
For example, in some petitioners with more than $1000 in medical debt in our study, that debt made up less than 5% of all unsecured debt. Instead, we chose a relative threshold of at least 20% of all unsecured debt. The 20% medical debt threshold leads to a conclusion that 5% of randomly selected petitioners and 26% of injured petitioners have bankruptcy with substantial medical debt. The randomly selected debtors represent the average bankruptcy petitioner; their medical debts may be the result of serious illness, or even TBI and SCI, treated at other medical centers.
To put these proportions in context, our previous study
found that the incidence of bankruptcy increased by 33% in the 5 years after TBI or SCI, suggesting that approximately 25% (ie, 0.33/1.33) of injured petitioners had bankruptcies that probably would not have occurred in the absence of injury.
It is likely that there is some overlap between the 26% of injured petitioners with substantial medical debt in this study and the estimated 25% of injured bankruptcy petitioners who would not have filed in the absence of injury.
We found substantial medical debt in patients with commercial insurance, including a few cases in which the entire acute care bill appeared as a debt on the bankruptcy petition. Possible explanations for this are that insurance status was incorrectly coded in the registry, patients misrepresented their insurance status to medical staff, the insurer denied coverage, or the bankruptcy petition was filed before the insurer had reimbursed the hospital. In 2004, Washington State repealed alcohol exclusion laws that allow insurance companies to deny coverage for alcohol- or narcotic-related injuries.
Washington State House of RepresentativesHouse Committee on Financial Institutions and Insurance An act relating to insurance coverage for injuries sustained because of alcohol or narcotic use.
This is particularly relevant given our finding that individuals with higher blood alcohol levels at the time of injury were more likely to have substantial medical debt at the time of bankruptcy. Other commercially insured petitioners may have been underinsured and unable to cope with cost-sharing strategies such as deductibles, copayments, coinsurance, or insurance caps.
A further possibility is that injured petitioners with employer-sponsored insurance were unable to return to work after injury and experienced a lapse in coverage leading to medical debts for rehabilitative care.
The high levels of medical debt in patients with Medicaid insurance at the time of injury show that medical debt is not purely an issue for those with commercial or no insurance. Recent work has shown how vulnerable the Medicaid population is to loss or disruption in eligibility. In Oregon State, Carlson et al
found that 31% of enrolled adults reported losing Medicaid coverage after the implementation of cost-saving strategies, and those with disruption in coverage were more likely to report medical debt. At the time of our study, the Washington State Medicaid program included a medically indigent provision to pay for the acute trauma care of low-income uninsured patients who did not quite meet the regular Medicaid income threshold.
Therefore, very few patients (n=2) were classified as uninsured in this study. However, several patients, classified as being insured with Medicaid in this analysis, may have become Medicaid ineligible soon after the acute care stay was over. The medically indigent program was discontinued in 2003.
Indirect Costs
There is strong evidence that TBI and SCI reduce the chances of long-term employment. Doctor et al
found that 42% of patients were unemployed 1 year after TBI, which was more than 4 times the expected rate. These outcomes may be worse for SCI; less than 15% of individuals report employment 1 year after injury.
and many households are ill prepared for sustained periods of unemployment. Therefore, the indirect costs of injury because of job loss or reduced earning power provide another plausible pathway between injury and financial insolvency. In the survey by Himmelstein et al,
21% of bankruptcy petitioners reported losing 2 weeks or more of income because of ill health.
Although we have no direct data on employment, we found that patients discharged to an SNF or rehabilitation unit were more likely to report income support from social security or other government programs. The monthly median household income was not markedly lower in the injured petitioners than in randomly selected petitioners. One interpretation of these observations is that some injured petitioners have been able to access government disability assistance, but this assistance has not come soon enough or has not kept income high enough to cope with pre- or postinjury debts. In addition, the finding that injured petitioners had fewer assets at the time of bankruptcy might signify that they had been forced to spend down their assets in an attempt to cope with medical debts and reduced income. However, another plausible explanation is that injury often affects relatively young individuals who have few assets and low income before, rather than as a consequence of, the injury.
Nonmedical Direct Costs
Our analysis only considers medical debt incurred through care received by the debtor and their dependents. In a few petitions, it became evident that a major factor in the bankruptcy was a large debt because of a personal injury claim by a third party who had been injured by the bankruptcy petitioner. The liability claim could include the medical expenses of the third party along with other expenses such as property damage. Because of the difficulty of attributing these liability claims, we did not categorize them as medical debts. Future work should explore the prevalence of postinjury bankruptcy because of large liability claims. We were also unable to fully explore other nonmedical factors, such as childcare and travel costs or family breakup,
caused by the injury that might increase costs or reduce income and lead to bankruptcy. However, our finding of high rates of divorce and separation among injured petitioners who were discharged to an SNF or rehabilitation unit suggests that more detailed analysis of these factors is important.
Strengths and Limitations
This is the first study that we are aware of that examined the prevalence of substantial medical debt in the bankruptcy petitions of people known to have had serious health problems. Unlike previous work that has relied on self-report, we used a hospital database to estimate the severity of the medical problems.
We believe that our data provide a conservative estimate of the amount of medical debt in the bankruptcies of patients with TBI and SCI. We included petitions filed up to 5 years after injury, by which time some medical debts have been paid via second mortgages, credit cards payments, or referred to generic debt-collection agencies. Some medical debts may not appear on the petition because the individual intends to repay them. Bankruptcy petitioners must state the source of debt on the petition; however, petitioners may be unable to recall the original sources, particularly if the debts were small.
Our data did not allow us to compare or control well for the severity of the TBI or SCI. Although the ISS is a good predictor of mortality in trauma patients as a whole, it may not adequately describe the long-term severity of an SCI or TBI. In addition, the modified FIM, assessed during the acute care stay, may not reflect a patient's long-term functional status. Patients who were discharged to an SNF or rehabilitation unit had similar levels of medical debt on the bankruptcy petition as patients discharged to their own home after acute care. This suggests that the amount of rehabilitative care needed is not strongly associated with the level of medical debt at bankruptcy. However, discharge location may be influenced by factors, for instance insurance status, besides medical need for rehabilitative care. Therefore, a larger sample size with better measures of long-term injury severity is needed to explore this further.
By focusing on bankruptcy petitions, we cannot provide information on other injured patients who default on medical debts or struggle to repay them without filing for bankruptcy. Our study took place in 1 district before substantive changes to the bankruptcy law in 2005.
but the impact of the tightening of bankruptcy laws after 2005 on bankruptcies with substantial medical debt is unknown.
The relatively small sample size is also a limitation of this study. We found important differences between injured and randomly selected petitioners in terms of assets, government assistance, and medical debt. However, the detection of more subtle differences, particularly between TBI and SCI patients, would require a larger sample size.
Conclusions
The number of under- and uninsured individuals in the U.S. is projected to increase.
Our study shows that some people are unable to access sufficient financial support to deal with overwhelming medical bills and eventually discharge their debts in the bankruptcy courts. There are many potential policy responses to this problem, and opinions on the best approach will vary.
Some will favor no or only minor adjustments to the current fragmented system of medical insurance. Examples include mandatory personal injury protection auto insurance and the promotion of low-premium, high-deductible medical insurance to protect the currently uninsured against catastrophic medical bills.
Emergency department use and subsequent hospitalizations among members of a high-deductible health plan [published erratum appears in JAMA 2008;299:171].
Others will argue that these piecemeal responses have not provided durable solutions in the past and that more radical changes, such as a single-payer national medical insurance, are needed to prevent high medical bills leading to financial distress and bankruptcy after injury.
aSPSS, 233 S. Wacker Dr, 11th Fl, Chicago, IL 60606.
Acknowledgments
We thank Mike Clark, BSc, Curtis Udy, and Margaret Hoyle at the Western District of Washington US Bankruptcy Court for their assistance in linking to the Western Washington CM/ECF database.
Factors associated with higher levels of injury severity in occupants of motor vehicles that were severely damaged in traffic crashes in Kentucky, 2000-2001.
Emergency department use and subsequent hospitalizations among members of a high-deductible health plan [published erratum appears in JAMA 2008;299:171].
Supported by the Royalty Research Fund of the University of Washington and the intramural research program of the National Institutes of Health.
No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.
Includes 1 patient who discharged themselves against medical advice.
01633-X&id=gr1.jpg){kind=link}
01633-X&id=gr2.jpg){kind=link}