 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
a Department of Pediatric Surgery
b Office of Clinical Sciences, Columbus Children's Hospital, Columbus, Ohio
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
METHODS. A retrospective analysis was performed for all children between the ages of 2 and 20 years who were treated for appendicitis between January 1, 2001, and December 31, 2003. Demographic variables included patient age, gender, race, insurance status, parental educational status, and income level. Coding data were used to identify patients with perforated appendicitis. The use of radiologic imaging was also analyzed.
RESULTS. During the 3-year period, 788 patients were treated for appendicitis. The racial distribution (white: 81%; black: 12%; other: 7%) was consistent with the demographic composition of the local population. The overall perforation rate was 25%, and the rate was significantly greater in the age group of <6 years, compared with older children. However, there were no significant differences in the perforation rate with respect to race, insurance status, educational level, or income status. Rates of radiologic imaging use were similar among all racial and socioeconomic groups.
CONCLUSIONS. Although racial and socioeconomic disparities in the rates of perforated appendicitis among children have been reported, we found no significant evidence for such inequality at our institution. This may reflect improved access, early diagnosis, and referral by primary care physicians in the community. Pooled national and multiple-state administrative databases have been used to highlight persistent disparities in health care. This study illustrates how single-institution data sources can be used to test a local hypothesis generated by national data, with surprisingly different results.
Key Words: appendicitis • outcome • racial disparity • perforation • children
Abbreviations: ICD-9—International Classification of Diseases, Ninth Revision • CT—computed tomography
The negative impact of lower socioeconomic status on health care outcomes is well known. Numerous studies also document the significant health care gaps between ethnic and racial groups even when access-related factors, such as income and insurance status, are controlled. Evidence for the problem of unequal access to health care was highlighted in the landmark report by the Institute of Medicine.1 Although the data used in the Institute of Medicine report were derived almost exclusively from adult patients, there is increasing evidence indicating significant disparities in selected areas of health care among children across different ethnic and racial groups, independent of socioeconomic status.2–6 This includes recent findings that certain racial, ethnic, and immigrant groups are at increased risk of perforated appendicitis.7–9
Acute appendicitis is considered an ideal subject of investigation regarding health care disparities because it is characterized by a consistent natural history, a single definitive treatment (appendectomy), a distinct negative outcome (perforation), and an absence of known biological predisposition to perforation in any racial or ethnic group.7,10 Indeed, perforated appendicitis is considered to be a sentinel marker for access to emergency care for time-dependent conditions.7,11,12 Therefore, a thorough understanding of factors contributing to higher rates of perforated appendicitis is crucial in the effort to promote equal access to health care.
Socioeconomic factors, such as lower incomes, lack of insurance, or public payer status, may be associated with increased perforation rates by reducing timely access to hospital care.12–18 At the inpatient hospital level, diagnostic delay and inappropriate initial care lead to increased rates of perforation.19,20 For disadvantaged minority populations, the effects of race and ethnicity may be additive with lower socioeconomic status in reducing access to timely care.
Many published studies demonstrating racial and ethnic disparities in appendicitis and other health care issues have relied on large administrative databases accrued from multiple institutions, regions, and states. Because of well-known methodologic problems with pooled administrative data, we embarked on this analysis to determine whether there are ethnic disparities in the assessment and treatment of children presenting with acute appendicitis in our institution.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The outcome of interest was complicated appendicitis, which included cases associated with perforation, gangrene, or peritonitis (ICD-9 diagnosis code 540.0 or 569.5), peritoneal or intestinal abscess (code 540.1 or 569.5), acute, chronic, or unspecified parametritis or pelvic cellulitis (code 614.3 or 614.4), other cellulitis or abscess, trunk (code 682.2), or other postoperative infection (code 998.59). This coding is consistent with that used in other administrative data-based appendectomy research.7,8,21 As in other studies, for simplicity we refer to all cases meeting these criteria as "perforated."
Independent variables extracted from the administrative data included patient age (<5, 5–12, or 13–20 years), race (white, black, or other), and payer (commercial, Medicaid, or other). With each patient's zip code, US Census data22 were used to estimate household income and parents' educational attainment. Zip code median incomes were categorized as high (<5% below the poverty level), moderate (5–15% below the poverty level), or low (>15% below the poverty level), and zip code median educational achievements were categorized as low (<25% post-high school education), moderate (25–50% post-high school education), or high (>50% post-high school education).
One health care delivery variable of interest was the use of imaging. In our institution, the use of radiologic imaging, especially ultrasonography and computed tomography (CT), for the diagnosis of appendicitis is 
60%.23 Because radiologic imaging has become a vital component of care for these patients, we investigated whether its use varied among the different racial or socioeconomic groups. Using charge code data, we identified all appendectomy-related imaging received by patients. Use of imaging was then categorized as none, abdominal CT scans only, abdominal radiographs only, abdominal radiographs and CT scans, or other (including use of ultrasonography and fluoroscopy).
Analyses of univariate effects were tested with 
2 tests. A multivariate model testing the effect of a variable in the presence of potential confounders was developed with logistic regression. Statistical significance was assigned at P < .05.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Use of imaging was associated strongly with perforation. Overall, some form of radiologic imaging was conducted in 71% of cases. The imaging rate increased to 82% when the analysis was limited to children with perforated appendicitis. Patients with perforated appendicitis were also more likely than patients with nonperforated appendicitis to undergo multiple types of imaging. More than 1 type of imaging was performed in 28.6% of nonperforated cases, compared with 43% of perforated cases. No relationship was found between the use of imaging and race (2 = 8.88, P = .35) (Table 2).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Whether inequalities exist in the care of children with appendicitis is an important question because perforated appendicitis is considered to be a sentinel marker for access to emergency care for time-dependent conditions.7,11,12 Therefore, a higher rate of appendiceal perforation for a particular demographic group could indicate inferior access to care, which might have implications for health policy. Studies that report large disparities based on race and socioeconomic factors often are those that use large administrative databases that reflect a larger population base. Pooled data, however, obscure regional differences in outcomes. For example, the recent report by Ponsky et al8 that showed significant racial disparities in the rates of appendiceal rupture included data from 30 institutions, including our hospital. Although data from the present study showed no effect of race, socioeconomic status, or insurance coverage, disparities in care in other communities may exist to a much greater degree, affecting the pooled results with large databases. However, such extreme outcomes can be revealed only through discrete analyses of regional data.
Other factors associated with higher rates of perforation in this study were younger age and use of >1 type of imaging. The finding of a higher rate of perforation (41.9%) among children <6 years of age is a well-established characteristic of this disease. Even when racial or other disparities were detected in previous studies, higher perforation rates among younger children were consistent findings.7,8,13,21 In addition, we were not surprised at the finding that higher perforation rates were associated with greater use of radiologic imaging. For the past 5 years, an appendicitis clinical pathway that advocates the use of ultrasonography or CT when perforation is suspected has been implemented in our institution. Detection of perforation, especially in association with intraabdominal abscess, has allowed us to use an initial nonoperative approach followed by interval appendectomy for a selected group of patients. With this protocol, we reported recently that 60% of children being evaluated for appendicitis in our institution underwent either ultrasonography or CT.23 A recent survey of North American pediatric surgeons also found that >90% of respondents routinely obtained either ultrasound or CT scans when there was suspicion of perforated appendicitis.30
One major limitation of this study was that the sample size analyzed did not provide adequate power to detect a difference of the magnitude found here. The sample size in our study was sufficient only to detect a black/white perforation rate difference of 15%, rather than the 6.4% found, assuming power of 0.80 and significance of .05. Because cases among white patients are accrued at a rate of nearly 7:1, compared with cases among black patients, a very long follow-up period (in which patterns of care might not be stable) would be required to obtain the power to detect small differences. Therefore, our lack of significance in racial disparity might have been in part a result of lack of power. This is likely the reason why most researchers have focused on the use of large, multiple-institution, administrative databases for disparity studies. A second major limitation of this study was that determination of perforation was based on hospital discharge coding data, which in turn relied on the surgeon's diagnosis at the time of surgery. The accuracy of our data would have been improved with confirmation of the final histopathologic diagnoses. This limitation applies to all studies that depend on the analysis of single-institution or pooled administrative databases and was highlighted in a recent report.31 A third limitation was that racial identification of patients in our database was somewhat problematic. Hospital admitting clerks are trained to ask patients directly about their race; however, we understand that, because of the sensitive nature of the question, some clerks guess race on the basis of appearance or surname. The extent and direction of misclassification at our institution are not known but we think they are small, because our study's racial distribution was consistent with that of our primary service area. Finally, our finding of no racial or socioeconomic disparity in the outcomes for childhood appendicitis may not reflect the situation in other communities. The racial distribution (white: 81%; black: 12%; other: 7%) of our patients is consistent with the demographic composition of the local population, which in turn is similar to that of the United States as a whole. Our institution is the only major children's hospital in central Ohio and has one of the largest pediatric primary care networks in the United States, with 1 primary care center on the hospital campus and 8 off-site "close to home" physician care centers. These close to home centers have been strategically located to extend the point of care to traditionally disadvantaged communities. Referral for pediatric surgical care is facilitated by direct communication between all regional pediatricians and an on-call pediatric surgeon on a 24-hour basis. The relatively easy access to specialist pediatric surgical care for all children, irrespective of insurance status and socioeconomic level, may be partly responsible for the absence of racial or socioeconomic disparity found in this study. Similar comprehensive centers exist throughout the nation, and their data may be similar to ours.
This study highlights the importance of recognizing that, although national data may suggest a broad trend toward disparate care for a given condition, the unique characteristics of any individual institution or system (eg, referral network, patient population, or on-going quality activities) may yield quite different results. Unfortunately, the cost associated with performing local analyses often prevents providers from conducting their own research, leaving facilities with the options of acting on national data, which may not apply to them, or not acting at all. This project illustrates how single-institution data sources can be used, relatively inexpensively, to test a local hypothesis generated by national data. Although these data do not contain all of the clinical detail required to describe the intricacy of clinical decision-making, they do serve to identify areas in which to focus research and areas in which no problem seems to exist. The usefulness of any administrative data system is limited by the quality, completeness, comprehensiveness, and accessibility of the available data and the availability of information systems staff members with the time and skills required to complete analyses such as those presented here.
An analysis such as this is not without cost. Databases must be created or linked if a comprehensive data warehouse is not already in place, and appropriately skilled staff members must be hired or diverted from other projects. Each institution must consider its own strategic, financial, and quality goals and set analytic priorities appropriately. We think that conducting local analyses is most cost-efficient when the result can be used to inform or to target major quality improvement efforts that require large investments of time and money. In such cases, the cost of conducting local analyses may be offset by the cost savings associated with implementing a more appropriately targeted program.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address correspondence to Benedict C. Nwomeh, MD, Department of Pediatric Surgery, Columbus Children's Hospital, 700 Children's Dr, Suite ED379, Columbus, OH 43205. E-mail: nwomehbe{at}chi.osu.edu
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  H. G. Herrod and C. F. Chang Potentially Avoidable Pediatric Hospitalizations as Defined by the Agency for Healthcare Research and Quality: What Do They Tell Us About Disparities in Child Health? Clinical Pediatrics, March 1, 2008; 47(2): 128 - 136. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2006 American Academy of Pediatrics. All rights reserved. |
||
|
||
