Effects of Race and Insurance on Outcomes of the Open Abdomen

Recent studies have suggested improved outcomes in surgical patients with healthcare insurance, whereas several others have noted disparities in access to health care, the care provided, and the aftercare of uninsured patients. Several different strategies exist in the management and prevention of the open abdomen secondary to abdominal compartment syndrome. To date, no study has evaluated the effects of race and insurance in patients with an open abdomen (OA). A retrospective review from our OA database was queried. All patients with an OA from January 2002 to December 2010 were included for analysis. Data analyzed included patients' demographics, race, insurance status, hospital charges, Injury Severity Scores, and outcomes. Insured patients were identified and compared with their uninsured counterparts. A total of 720 patients were treated for an OA during the study period. Of these, 273 (37.9%) died within their hospital stay. Patients who died were noted to be older and sicker with higher Acute Physiology and Chronic Health Evaluation (APACHE) II and Simplified Acute Physiologic Scores (27.6 vs 18.2, P<0.001 and 54.6 vs 38.5, P<0.001, respectively). Logistic regression analysis revealed that age, APACHE II, and Injury Severity Scores were independently associated with mortality. From our categorical variables, race was not associated with worse outcomes. In addition, being uninsured was significantly associated with increased mortality (odds ratio, 1.67; 95% confidence interval, 1.1 to 2.6; P = 0.05). ''Self-pay'' status was associated with increased mortality even after adjusting for severity of illness. Further studies incorporating baseline comorbidities need to be undertaken to further assess the reasons for these disparities.

IN 2009, THE United States showed a drastic rise in the uninsured to 50.7 million individuals, whereas the number of insured persons dropped for the first time in 20 years.1 This, coupled with the U.S. notoriety as having the worse coverage for its citizens as an industrialized nation, has been a matter of national debate for some time now.2

A lack of health insurance leads to lack of adequate healthcare screening, delay in diagnosis, and more undiagnosed diseases.3-7 There is even some concern that uninsured patients receive different treatment than those with insurance.7 Several studies have shown a direct relationship between lack of insurance and mortality.2, 8, 9 In 2002, the Institute of Medicine estimated that 18,314 Americans between the ages of 18 and 64 years died annually from lack of insurance.10 Wilpur et al.8 later noted that this was based on historical data and recalculated this figure to 35,327 deaths annually.

Within the traumatically injured population, some studies revealed that uninsured patients demonstrated worse outcomes when compared with their counterparts. 9, 11-13 Multiple laws and protocols have been enacted to ensure that equal care is provided regardless of payer status within this population. Despite these measures, the uninsured still appear to be at increased risk of death. Several reasons for this have been hypothesized including delay of treatment, administration of different care, and a lower rate of health literacy.9

The development of abdominal compartment syndrome (ACS) in the critically ill patient is a significant source for morbidity and mortality. Its management includes decompression with a subsequent open abdomen (OA) when all other measures fail; in addition, the traumatically injured patient requiring damage control laparotomy also requires temporary OA until correction of instability.14-16

The overall treatment of the OA is continuously evolving and is aimed at correcting the underlying condition, including providing the patient adequate nutrition, avoiding fixity, and attempting an early closure of the abdomen.14-16 Despite this, most treatments are based on hospital resources and surgeon experience.

To date, no study has evaluated the association of race and insurance in patients with OA. At our institution, we have a large database of patients who have been treated for OA. Therefore, we attempted to investigate the role of insurance status and race in patients with OA. We hypothesized that there would be a disparity in mortality among the uninsured patients.

Methods

Data were retrospectively reviewed from a prospectively gathered single-institution OA database. All patients treated for OA were reviewed from January 2002 to December 2010 in an urban Level I trauma center. As a result of its retrospective nature, no patients were excluded. Data analyzed included demographics, race, insurance status, hospital charges, hospital length of stay (LOS), intensive care unit (ICU) LOS, and inhospital mortality. Severity of illness scores that were evaluated included Acute Physiology and Chronic Health Evaluation (APACHE) II and Simplified Acute Physiology Score (SAPS) 2. For trauma patients, the Injury Severity Scores (ISS) were likewise evaluated.

Race was broken down into the following categories: white, black, Hispanic, and Asian. Any patients who did not fit into these main descriptions were listed as ''other.'' Payer status was listed as follows: commercial insurance, managed care,Medicaid,Medicare, and selfpay. Payer status was ultimately identified at the end of a patient's hospital stay because any insurance (or lack thereof) before admission could not be delineated.

We performed a subgroup analysis of two different groups. The first group was trauma-only patients, which in theory should be a younger group with less preexisting comorbidities. Likewise, we performed an analysis of the nontrauma patients, which comprised the second group. As before, these groups were broken down among race and payer status.

Statistical analysis was performed using SPSS software, Version 17 (IBM, SPSS Statistics, Chicago, IL). Student's t test and x2 analysis were used for continuous and categorical data, respectively. A stepwise logistical regression analysis was performed to determine independent risk factors for mortality. A P value of 0.05 was considered statistically significant.

Results

During the study period, 720 patients were treated for OA. Of these, 472 patients were initially admitted secondary to trauma and 248 were admitted for diagnosis other than trauma. The mean age was 43.6 years. The total number of male and female patients were 522 (72.5%) and 198 (27.5%), respectively. The mean hospital LOS and ICU LOS were 28 and 12.7 days, respectively. Overall mortality was 38 per cent and number days for an OA were 8 ± 8. Overall demographical data is seen in Table 1. Patients who died were significantly older and sicker based on their severity of illness scores. Patients who survived were noted to have both longer ICU and hospital LOS. Survivors were also likely to have higher hospital charges (Table 1).

When patients were categorized by race, the majority of patients were white (53%) followed by black (23%) and Hispanic (21%), respectively (Table 2). Asians and those listed as ''other'' comprised only 2.6 per cent of the population. Six patients (1%) had no race recorded within their chart. When overall mortality was compared among races, blacks and whites showed a trend toward survival; however, this was not statistically significant (Table 2).

Of the 720 patients, only 598 (83%) had a payer status recorded. The majority of this subgroup had commercial insurance (50%). The remaining payers were fairly evenly distributed with self-pay and managed care being the next most prominent payer status at 15 and 14 per cent, respectively. Medicaid and Medicare were the least prominent payer status at 12 and 9 per cent, respectively (Table 3). When payer status was separated into trauma versus nontrauma, the nontrauma population showed a noteworthy increase in the percentage of Medicare patients with a subsequent drop in both commercial and self-pay insurance (Fig. 1).

Trauma patients had a mortality of 32 per cent versus nontrauma patients whose mortality reached 49 per cent. This was noted to be statistically significant (P < 0.001) (Table 4). Patients were also noted to be younger in the trauma population (median 34 vs 57 years; P < 0.001) and more likely to be male (79 vs 59%; P < 0.001).

A multivariate analysis was performed to identify independent factors associated with increased mortality (Table 5). Age was still associated with increased mortality in the final analysis (odds ratio [OR], 1.04; 95% confidence interval [CI], 1.02 to 1.05; P < 0.001). Higher APACHE II (OR, 1.14; 95% CI, 1.11 to 1.19; P < 0.001) and ISS (OR, 1.04; 95% CI, 1.02 to 1.06; P < 0.001) scores were likewise associated with increased mortality. Although SAPS 2 scores showed increased mortality in the univariate analysis, it failed to show any association with mortality in the multivariate analysis (OR, 1.01; 95% CI, 098 to 1.04; P 4 0.40). No individual race was identified as being associated with increased mortality. ''Self-pay'' was the only payer status associated with mortality in the final analysis (OR, 2.49; 95% CI, 1.11 to 1.19; P < 0.001). Interestingly, this persisted even when patients were separated into nontrauma and trauma patients (Table 6).

Conclusion

A disparity among the uninsured is not a new topic.2-9, 11 An ever increasing number of studies have shown that the uninsured receive decreased access to health care with lack of adequate health screening and delay in diagnosis of diseases. More recently, several studies have attempted to look at trauma patients who, in theory, should receive identical treatment after arriving into the healthcare system. Haider et al.12 identified patients through the National Trauma Data Bank and noted that patients were at increased risk of mortality based on insurance status and race. The authors noted that not only did minority race have an increased risk of mortality, but that an ''uninsured'' status worsened the finding. Unfortunately, the authors failed to take into account other factors such as preexisting comorbidities. Rosen et al.9 performed a very similar study using the same database.When the authors took into account comorbidities, they found that race no longer played a role in mortality and that only insurance status played a major role in mortality prediction.

Erickson et al.18 attempted to look at disparities within the ICU. The authors noted that once again there were no differences in mortality based on ethnicity once they adjusted for APACHE admission scores. They did note that black patients were more likely to have more acute physiologic derangements within the ICU but this did not translate over into increased mortality. Interestingly enough, the authors did not find an association with payer status and mortality either.

In our study, only ''self-pay'' status was associated with increased mortality in patients within the OA population. This prominent increase in mortality persisted even after adjusting for severity of illness for the patient population. There are several possible reasons for this postulated by Rosen et al.:

1. The uninsured may undergo a delay in treatment;

2. The uninsured may receive different levels of care than the insured; and

3. The uninsured may possess lower rates of health literacy, which leads to poor communication with the physician.9

Although it is impossible to determine whether there was either a delay in treatment or decreased health literacy within our database, we know that all the patients receive identical care after their abdomens have been leftopen. Our database is composed of patients who are managed by the acute care service at our institution and all follow evidence-based medicine guidelines that were created by the service. If we exclude treatment variance, then this implies that the uninsured patient likely has undiagnosed preexisting comorbidities that lead to worse outcomes. Lack of access to adequate health care and lack of proper healthcare screening by the uninsured population could explain this phenomenon.3-6, 19 Unfortunately, we are unable to garner this information from our database.

It was also noted that trauma patients with OA fared better than the nontrauma patient. This is likely explained by the different nature of the patients. It is not unusual for trauma patients who undergo emergent exploratory laparotomies to be leftopen. This population also tends to be younger, which would lead to less comorbidities. The nontrauma population, as demonstrated in our study, tends to be older and also included more patients that developed ACS for medical admissions. This would imply the patients are likely to be sicker with more preexisting comorbidities.

There are a few weaknesses to our study. As a result of the retrospective nature of the study, we are unable to gather baseline comorbidities that are not included within the database. Also, patient's payer status is determined at the time of discharge. Therefore, many patients who were uninsured at the time of admission may have received Medicaid during their hospital stay. So if an uninsured patient survived long enough within the hospital stay, they may have been moved over to the Medicaid population, thereby falsely elevating the overall mortality within the ''self-pay'' group. Unfortunately, there is no way to identify patients' admission payer status within our hospital system. Lastly, although our institution currently has the largest database for OA, it may be too underpowered to detect differences within minorities.

We have demonstrated that ''self-pay'' status leads to increased mortality within patients with OA. This is not likely the result of lacking insurance but more likely related to increased morbidities that were previously undetected. Future studies incorporating baseline comorbidities need to be undertaken to evaluate this effect of uninsured patients.

REFERENCES

1. US Census. Income, Poverty, and Health Insurance Coverage in the United States: 2009. Available at: www.census.gov. Accessed November 2010.

2. Wilper AP,Woolhandler S, Lasser KE, et al. Health insurance and mortality in US adults. Am J Public Health 2009;99:2289-95.

3. Kleinman JC, Gold M, Makuc D. Use of ambulatory medical care by the poor: another look at equity. Med Care 1981;19: 1011-29.

4. Ayanian JZ, Wiessman JS, Schneider EC, et al. Unmet health needs of uninsured adults in the United States. JAMA 2000;284: 20261-9.

5. Ayanian JZ, Zaslavsky AM,Weissman JS, et al. Undiagnosed hypertension and hypercholesterolemia among uninsured and insured adults in the Third National Health and Nutrition Examination Survey. Am J Public Health 2003;93:2051-4.

6. De Bosset V, Atashili J,MillerW, Pignone M. Health Insurancerelated disparities in colorectal cancer screening in Virginia. Cancer Epidemiol Biomarkers Prev 2008;17:834-7.

7. Cram P, Pham HH, Bayman L, Vaughan-Sarrazin MS. Insurance status of patients admitted to specialty cardiac and competing general hospitals: are accusations of cherry picking justified? Med Care 2008;46:467-75.

8. Wilper AP, Woolhanler S, Lasser KE, et al. Health Insurance and Mortality in US Adults. Am J Public Health 2008; 99:2289-95.

9. Rosen H, Saleh F, Lipsitz S, et al. Downwardly mobile. The accidental cost of being uninsured. Arch Surg 2009;144:1006-11.

10. Institute of Medicine. Care without Coverage, Too Little, Too Late. Washington, DC: The National Academies Press; 2002.

11. Herring AA, Woolhandler S, Himmelstein DU. Insurance status of US organ donors and transplant recipients: the uninsured give, but rarely receive. Int J Health Serv 2008;38:641-52.

12. Haider AH, Change DC, Efron ET, et al. Race and insurance status as risk factors for trauma mortality. Arch Surg 2008;143: 945-9.

13. Greene WR, Oyetunji TA, Bowers U, et al. Insurance status is a potent predictor of outcomes in both blunt and penetrating trauma. Am J Surg 2010;199:554-7.

14. Cheatham ML, Safcsak K. Is the evolving management of intra-abdominal hypertension and abdominal compartment syndrome improving survival? Crit Care Med 2010;38:402-7.

15. Cheatham ML, Safcsak K. Skin-only closure is superior to skin grafting in the open abdomen. Acta Clin Belg 2007;62:262.

16. Bjorck M, Bruhin A, Cheatham ML, et al. Classification- important step to improve management of patients with an open abdomen. World J Surg 2009;33:1154-7.

17. Haas JS, Goldman L. Acutely injured patients with trauma in Massachusetts: differences in care and mortality, by insurance status. Am J Public Health 1994;84:1605-8.

18. Erickson SE, Vasilevskis EE, Kuzniewicz MW, et al. The effect of race and ethnicity on outcomes among patients in the intensive care unit: a comprehensive study involving socioeconomic status and resuscitation preferences. Crit Care Med 2011;39: 429-35.

19. Kong AL, Yen TW, Pezzin LE, et al. Socioeconomic and racial differences in treatment for breast cancer at a low-volume hospital. Ann Surg Oncol 2011;18:3220-7.

JASON M. CLARK, M.D., MICHAEL L. CHEATHAM, M.D., KAREN SAFCSAK, R.N., RODRIGO F. ALBAN, M.D.

From Orlando Health-Surgical Education, Orlando, Florida

Oral presentation at the Florida Chapter Meeting of the American College of Surgeons, February 2011; and poster presentation at the Fifth World Congress on Abdominal Compartment Syndrome, August 2011 in Orlando, Florida.

Address correspondence and reprint requests to Jason Michael Clark, M.D., Surgical Resident, Medical Education, Orlando Health-Surgical Education, 86W. Underwood, Suite 201, Orlando, FL 32806. E-mail: [email protected].