Factors associated with length of stay and hospital charges for patients hospitalized with mouth cellulitis

Transcription

1 Factors associated with length of stay and hospital charges for patients hospitalized with mouth cellulitis Min Kyeong Kim, BA, a Romesh P. Nalliah, BDS, b Min Kyeong Lee, DMD, c and Veerasathpurush Allareddy, BDS, MBA, MHA, MMSc, PhD, d Boston, Massachusetts HARVARD SCHOOL OF DENTAL MEDICINE Vol. 113 No. 1 January 2012 Objective. Dental conditions that are neglected could progress to infectious lesions that are severe enough to require hospital admission for treatment. The objective of this study was to examine outcomes in patients hospitalized for cellulitis and abscesses of mouth in the USA in year Study design. The nationwide inpatient sample for the year 2008, a component database of the Healthcare Cost and Utilization Project, was used for the current study. All hospital discharges with a primary diagnosis for cellulitis or abscess of mouth (ICD-9-CM code 528.3) were selected for analysis. Outcomes, including length of stay in hospital (LOS) and hospitalization charges, were computed from the database and projected to national levels by using the discharge weight variable. The predictor variables were composed of sets of heterogeneous variables grouped into the following categories: demographic, health-related (comorbid conditions), hospital-specific, and insurance-related. The primary outcome variables were LOS and hospital charges. Multivariable linear regression analysis models were used to examine the association between predictor and outcome variables. Results. A total of 4,044 hospital discharges were attributed primarily to cellulitis or abscess of mouth. About 45% of these discharges occurred in those aged between 18 and 45 years. The mean length of stay in hospital was 3.9 days, and the mean hospital charge was $24,290. The total USA hospitalization charge was close to $98 million. Private insurance plans were the major payers, accounting for $31 million of hospitalization charges. About 88% of all hospitalization were discharged routinely after treatment, and 2% were transfered to another short-term hospital. Conclusions. This study examines outcomes in patients hospitalized for cellulitis or abscess of mouth. Future studies must focus on identifying cohorts that are more prone to developing odontogenic infections that are severe enough to warrant hospitalization. (Oral Surg Oral Med Oral Pathol Oral Radiol 2012;113:21-28) Cellulitis is a common skin infection that manifests into a diffuse inflammation in the aponeurotic spaces. 1,2 In particular, when several anatomic spaces become involved from a spread of infection in the mouth, cellulitis of the mouth can result. Even though mild-tomoderate cellulitis of the mouth can be treated in outpatient settings, delayed treatments can result in more severe spread of infection owing to the complexity of head and neck anatomy. 3 If left untreated, dental conditions such as localized odontogenic infections, which is considered to be one of the major etiologies for cellulitis of the mouth, 4,5 can spread through various facial planes and compromise multiple compartments after perforation of its adjacent bone. Epidemiologic review of a DMD Student. b Senior Tutor and Instructor, Department of Restorative Dentistry and Biomaterials Sciences. c Orthodontic Resident, Department of Developmental Biology. d Instructor, Department of Developmental Biology. Received for publication Dec 4, 2010; returned for revision Dec 27, 2010; accepted for publication Jan 8, Elsevier Inc. All rights reserved /$ - see front matter doi: /j.tripleo facial infections in hospitalized pediatric patients demonstrated that 50% of total hospital facial infections were due to odontogenic cellulitis. 6 There is a paucity of knowledge regarding epidemiology of hospitalization outcomes in patients with cellulitis using nationally representative data. For example, when an infection involving the floor of the mouth spreads to the submandibular, sublingual, and submental spaces, a cellulitis known as Ludwig angina can manifest. 7 Described by Willhelm Frederick von Ludwig in 1836, Ludwig angina is a medical emergency that rapidly progresses through the floor of the mouth into the parapharyngeal space. One case report described a 25-year-old man who suffered from a sudden death from Ludwig angina secondary to periodontal abscesses associated with lower molars. 8 Another case report discussed a 25-year-old healthy man with multiple carious teeth and noted periapical pathology, who died of a massive hemorrhage resulting from maxillofacial infection. 9 Urgent hospitalization and timely management are necessary to prevent possible asphyxiation and death. Although fatal outcomes, such as Ludwig angina, are rare, rapid spread of dental infections regarding teeth 21

2 OOOO 22 Kim et al. January 2012 and/or associated oral tissues can lead to other lifethreatening complications, including mediastinitis, septic shock, and massive hemorrhage, 9,10 that can lead to poor hospitalization outcomes. Therefore, dental conditions that are neglected could progress to infectious lesions that are severe enough to warrant hospital admission for treatment. In a 5-year retrospective study of pediatric emergency department visits, Ladrillo et al. discovered that there was a substantial increase in the inpatient admission of patients with dental problems during thestudy period (1997 through 2001). 11 In a 9-year retrospective study done in Britain, Moles and Ashley found that there was an ongoing increase in hospital admissions due to dental conditions in children and adolescents. 12 Although such findings emphasize an association between overlooked dental conditions and subsequent hospitalization, the results cannot be generalized to the national level in the USA. Cellulitis of the mouth is usually a preventable dental condition; therefore, identification of high-risk groups and early interventions, including proper diagnosis and timely delivery of adequate dental care, can alter the need for hospital admission and its outcome. 13 However, our current knowledge in hospital outcomes associated with cellulitis of the mouth is based only on single-center studies. 1,11,14 There is lack of literature examining costs incurred in hospitals for treating cellulitis and the role of associated comorbid conditions that may contribute to the overall prognosis of cellulitis. We hypothesized that there could be certain patient factors (including demographic, health-related, such as comorbid conditions, and insurance-related) or hospital factors that may be associated with hospitalization outcomes, including length of stay and hospital charges. Some of these factors could be potentially modifiable to enhance outcomes. The objectives of the present study were to examine outcomes in patients hospitalized for cellulitis of the mouth and to identify possible patientand hospital-level factors associated with such outcomes. MATERIAL AND METHODS Data source and patient selection This study is a retrospective analysis of a nationwide hospital discharge dataset. The nationwide inpatient sample (NIS) of the Healthcare Cost and Utilization Project (HCUP) for the year 2008, provided by the Agency for Healthcare Research and Quality (AHRQ), Rockville, Maryland, for research purpose, was used for the present study. 15 The NIS, a 20% stratified probability sample of all nonfederal acute-care general hospitals in the USA, is the largest such database, covering 8 million hospitalizations drawn from 1,056 hospitals in 42 states. Institutional Review Board approval was not required, because this study was a secondary data analysis of an NIS dataset made available by AHRQ. One of the authors obtained data from AHRQ after completing the data user agreement with HCUP and conducted all data analyses. All hospital discharges with a primary diagnosis for cellulitis (International Classification of Diseases, 9th Revision, Clinical Modification 16 [ICD- 9-CM] diagnosis code 528.3) were selected for analysis. Outcomes, including length of stay in hospital and hospitalization charges, were computed from the database and projected to national levels by using the discharge weight variable. Outcome variables Two outcome variables, hospital charges and length of stay, were obtained from the dataset and examined in the current study. All hospital charges are in 2008 dollars. Length of stay indicates the number of days that a patient stayed in the hospital since admission. Predictor variables The predictor variables were composed of a heterogeneous set of patient- and hospital-related variables. The demographic variables included age, gender, and insurance status (Medicare, Medicaid, private insurance [including Blue Cross, commercial carriers, private health maintenance organizations, and preferred provider organizations], uninsured [self-pay or no charge], and other insurance plans [including workers compensation, Civilian Health and Medical Program of the Uniformed Services, Civilian Health and Medical Program of the Department of Veterans Affairs, Title V, and other government programs]). Health status related variables included presence of comorbid conditions. We also obtained information regarding hospital-related factors, such as teaching status (teaching or nonteaching), location (rural or urban), geographic region (northeast, midwest, south, or west), and bed size (small, medium, or large number of beds), and included them in all analyses. 15 All predictor variables were obtained from the dataset. Age was used as a continuous variable, although all other demographic, health status related, and hospital related variables were used as categoric variables as provided in the dataset. Statistical analyses We used multivariable linear regression analysis to examine the association between hospital charges (outcome variable) and the predictor variables (age, gender, insurance status, presence of comorbid conditions, type of admission, and hospital characteristics). Because the hospital charge data were highly skewed, we log-transformed the data and used these log-transformed hospi-

3 OOOO ORIGINAL ARTICLE Volume 113, Number 1 Kim et al. 23 Table I. Descriptive statistics of the patient and hospital samples (n 4,044) Characteristic Response n (%) Gender Male 2,092 (52.0%) Female 1,928 (48.0%) Insurance status Medicare 807 (20.1%) Medicaid 764 (19.0%) Private 1,496 (37.2%) Uninsured 773 (19.2%) Others 184 (4.6%) Type of admission Elective 355 (8.8%) Emergency/urgent 3,683 (91.2%) Race White 2,060 (63.9%) Black 585 (18.1%) Hispanic 358 (11.1%) Asian or Pacific Islander 88 (2.7%) Native American DS Others 128 (4.0%) Bed size Small 470 (11.6%) Medium 874 (21.6%) Large 2,699 (66.7%) Location Rural 523 (12.9%) Urban 3,520 (87.1%) Teaching status Nonteaching 2,042 (50.5%) Teaching 2,002 (49.5%) Geographic region Northeast 679 (16.8%) Midwest 875 (21.6%) South 1,557 (38.5%) West 931 (23.0%) Comorbid condition AIDS 20 (.5%) Alcohol abuse 126 (3.1%) Deficiency anemias 324 (8.0%) Rheumatoid 82 (2.0%) arthritis/collagen vascular diseases Chronic blood loss anemia DS Congestive heart failure 95 (2.3%) Chronic pulmonary 470 (11.6%) disease Coagulopathy 44 (1.1%) Depression 259 (6.4%) Diabetes, uncomplicated 518 (12.8%) Diabetes with chronic 43 (1.1%) complications Drug abuse 126 (3.1%) Hypertension 1,202 (29.7%) Hypothyroidism 169 (4.2%) Liver disease 29 (0.5%) Lymphoma 19 (0.5%) Fluid and electrolyte 645 (15.7%) disorders Metastatic cancer 40 (1.0%) Neurological disorders 157 (3.9%) Obesity 327 (8.1%) Paralysis 25 (0.6%) Peripheral vascular 61 (1.5%) disorders Psychoses 138 (3.4%) Pulmonary circulation 160(.4%) disorders Renal failure 96 (2.4%) Solid tumor without metastasis 55 (1.4%) Table I. Continued Characteristic Response n (%) Peptic ulcer disease 0 (0%) excluding bleeding Valvular disease 47 (1.2%) Weight loss 97 (2.4%) Individual cell counts may not add up to the global cell counts because of missing values and the differences arising from variance computations when using the discharge weight variable. DS, Discharge information suppressed (individual cell counts for these data are 10, and the HCUP data user agreement precludes us from presenting these estimates, which were combined into one group and presented. tal charges as the outcome variable in the multivariable linear regression model. The NIS dataset provided information regarding comorbid conditions. 15 We tested each comorbid condition bivariately by using simple linear regression analysis and included in the final multivariable linear regression model all comorbid conditions that were associated significantly with hospital charges at P.05. The same analytical approach that was used for hospital charges was also used for modeling length of stay. In all analyses, we computed variances according to the Taylor linearization methods by using a with replacement design. The NIS hospital stratum was the stratification unit, and each hospitalization was the unit of analysis. We used the discharge weight variable that is assigned to each case to project to national estimates. In all of the multivariable regression models, we adjusted for the effects of clustering of outcomes within hospitals. All statistical tests were 2 sided, and we considered P.05 to be statistically significant. All statistical analyses were performed using Statistical Analysis Systems Software, version 9.2 (SAS Institute, Cary, NC), and Sudaan version (Sudaan, Research Triangle Park, NC) software. RESULTS A total of 4,044 hospital discharges were attributed primarily to cellulitis or abscess of the mouth in the USA during the year Characteristics of patients hospitalized for cellulitis are summarized in Table I. About 45% of these discharges occurred in those aged between 18 and 45 years. There was an approximately equal distribution of male (52%) and female (48%) patients. Private insurance was the primary payer (37.2% of all hospitalizations). A majority of the admissions occurred on an emergency and/or urgent basis (63.9% of all hospitalizations; Table I). The mean length of stay in hospital was 3.9 days, and the mean hospital charge was $24,290. The total USA hospitalization charge was close to $98 million. Inhospital mortality rate was 1% (15 patients out of

4 OOOO 24 Kim et al. January 2012 Table II. Factors associated with hospitalization charges Characteristic Estimate (95% CI) $ change from mean P value Age (1-year increase) ( ) $ Gender Female 0.01 ( ) $ Male Presence of comorbid conditions Deficiency anemias 0.29 ( ) $8, Chronic blood loss anemia 2.06 ( ) $165, Coagulopathy 0.39 ( ) $11, Diabetes, uncomplicated 0.20 ( ) $5, Drug abuse 0.40 ( ) $11, Hypertension 0.04 ( ) $ Fluid and electrolyte disorders 0.37 ( ) $10, Metastatic cancer 0.61 ( ) $20, Obesity 0.33 ( ) $9, Pulmonary circulation disorders 0.85 ( ) $32, Weight loss 0.99 ( ) $40, Type of admission Elective 0.03 ( ) $ Emergency/urgent Location of hospital Urban 0.54 ( ) $17, Rural Teaching status of hospital Teaching 0.05 ( ) $1, Nonteaching Hospital bed size Large no. of beds 0.23 ( ) $6, Small/medium no. of beds Geographic region North East 0.11 ( ) $2, Midwest 0.03 ( ) $ West 0.39 ( ) $11, South Insurance Medicare 0.02 ( ) $ Medicaid 0.15 ( ) $3, Uninsured 0.19 ( ) $5, Other insurance plans 0.31 ( ) $8, Private insurance CI, Confidence interval. total 4,044). Private insurance plans were the major payers, accounting for $31 million of hospitalization charges. About 88% of all hospitalizations were discharged routinely following treatment; 2% were transfered to another short-term hospital. Hypertension was prevalent in more than one-fourth of all hospitalizations (29.7%). Other commonly present comorbid conditions were fluid and electrolyte disorders (15.7%), uncomplicated diabetes (12.8%), chronic pulmonary disease (11.6%), and obesity (8.1%; Table I). Urban hospitals (87.1%) and large bed hospitals (66.7%) accounted for a majority of hospitalizations (Table I). Results on the association between hospital charges and the independent variables from the multivariable analysis are summarized in Table II. After adjustments for all patient- and hospital-level factors that were available from the dataset, patients with comorbid conditions, including deficiency anemia (P.01), chronic blood loss anemia (P.0001), uncomplicated diabetes (P.03), drug abuse (P.03), fluid and electrolyte disorders (P.0001), metastatic cancer (P.04), obesity (P.01), pulmonary circulation disorders (P.04), and weight loss (P.0001), were associated with significantly higher hospital charges than those without these comorbid conditions. Each 1-year increase in age was also associated with higher hospital charges (P.02). Patients covered by other insurance plans, including workers compensation, Civilian Health and Medical Program of the Department of Veterans Affairs, Title V, and other government programs, were associated with higher hospital charges compared with those covered by private insurance plans (P.02). The uninsured also had higher hospitalization charges compared with those covered by private insurance plans

5 OOOO ORIGINAL ARTICLE Volume 113, Number 1 Kim et al. 25 Table III. Factors associated with length of stay Characteristic Estimate (95% CI) Days change from mean P value Age (1-y increase) ( ) Gender Female 0.05 ( ) Male Presence of comorbid conditions Deficiency anemias 0.24 ( ) Chronic pulmonary disease 0.09 ( ) Coagulopathy 0.34 ( ) Diabetes, uncomplicated 0.19 ( ) Drug abuse 0.33 ( ) Hypertension 0.12 ( ) Lymphoma 0.02 ( ) Fluid and electrolyte disorders 0.32 ( ) Metastatic cancer 0.61 ( ) Neurologic disorders 0.27 ( ) Obesity 0.12 ( ) Psychoses 0.30 ( ) Solid tumor without metastasis 0.24 ( ) Weight loss 0.87 ( ) Type of admission Elective 0.03 ( ) Emergency/Urgent Location of hospital Urban ( ) Rural Teaching status of hospital Teaching 0.08 ( ) Nonteaching Hospital bed size Large no. of beds 0.06 ( ) Small/medium no. of beds Geographic region Northeast 0.06 ( ) Midwest 0.01 ( ) West 0.06 ( ) South Insurance Medicare 0.03 ( ).72 Medicaid 0.12 ( ) Uninsured 0.06 ( ) Other insurance plans 0.29 ( ) Private insurance (P.03). Hospitalizations occurring in urban hospitals were associated with higher hospital charges compared with those occurring in rural hospitals (P.0001). Hospitalizations occurring in large-bed-size hospitals were associated with significantly higher hospital charges compared with those in small/medium-bed-size hospitals (P.01). Results on the association between length of stay and the independent variables from the multivariable analysis are summarized in Table III. Patients with comorbid conditions, including deficiency anemia (P.01), uncomplicated diabetes (P.01), drug abuse (P.01), fluid and electrolyte (P.0001), metastatic cancer (P.01), neurologic disorders (P.01), psychoses (P.01), and weight loss (P.0001) were associated with significantly longer-length of stay than those without the presence of these comorbid conditions. Patients with other insurance plans, including workers compensation, Civilian Health and Medical Program of the Department of Veterans Affairs, Title V, and other government programs were associated with longer length of stay compared with those covered by private insurance plans (P.01). DISCUSSION The present study explores the prevalence of hospitalizations primarily resulting from cellulitis or abscess of the mouth. The results indicate that a considerable amount of hospital resources are being used to treat cellulitis in hospital settings. The presence of comorbid

6 OOOO 26 Kim et al. January 2012 conditions was associated with a significantly higher hospitalization charges and prolonged duration of stay in hospitals. It is likely that complications occurring during hospitalization, such as spread of infection and worsening of symptoms, result in poor hospitalization outcomes. Multiple measures at each step in the admission process can be taken to reduce the overall hospital resources. Before admission, adequate diagnostic measures should be taken to identify the high-risk groups among those with cellulitis or abscess of the mouth. Radiographs or computerized tomography can be used to detect a presence of foreign body, gas in tissue, and even osteomyelitis. 2 Further imaging modalities, such as magnetic resonance imaging or ultrasound, can be supportive in distinguishing cellulitis from an abscess, 16 which can direct providers in delivering adequate and timely treatments to eliminate the causative agents. Blood culture and proper screening panel can be used as a definitive test for identification of the causative factor for the infection as well. 2 Along with proper diagnostic measures, there is a need for standardized protocols and guidelines for inhospital treatment approaches and prescriptions to prevent worsening of the preexisting comorbid conditions. Depending on the severity of infection and the presence of comorbid conditions that may place the patients at more compromised state, surgical debridement and/or antibiotics can be chosen empirically to facilitate better prognosis of the infection with reduced hospital charges and length of stay. Flynn et al. demonstrated that there was an association between an increase in length of stay among the admitted patients with severe odontogenic infections and the number of involved anatomic spaces and severity of infection. 17 In a single-center study looking at patients with severe cervicomediastinal abscess of odontogenic origin, Kinzer et al. recommended that aggressive antimicrobial and surgical interventions, along with early diagnosis, are crucial in patient management and successful outcome. 14 Although untreated cellulitis of the mouth can result in almost 100% mortality from possible airway obstruction and sepsis, suitable treatments can significantly lower the mortality rate. 18 Furthermore, appropriate airway support in managing patients with severe soft tissue infections of head and neck can help to minimize morbidity and mortality. 19 In addition to the several measures that can be taken by a health care provider in admitting patients with abscess or cellulitis of the mouth to obtain a favorable outcome, dental care providers can perform a number of interventions. Because recurrent cellulitis is known to be more common in patients with histories of major surgery, such as coronary artery bypass graft with saphenous vein and radical mastectomy with axillary lymph node dissection, 2 a thorough screening of the medical history in patients who present with multiple carious teeth and inadequate oral hygiene by a dental care provider will allow a closer evaluation of those with odontogenic infection and thereby facilitate preventing future hospitalizations. A dental care provider should also become familiar with diverse features of odontogenic infections to identify high-risk patients that may require future hospitalization. Hwang et al. discuss the variation in clinical findings between hospitalized and nonhospitalized patients suffering from odontogenic infections and their potential use for identifying patients who may require hospitalization. 20 Being able to recognize such differences early on may help clinicians to correctly triage the patients. Because untreated odontogenic infections can manifest to abscess or cellulitis of the mouth, both patient education on regular dental visits and government policy on providing better access to dental health care to the public need to be addressed. Some of the existing barriers to dental care, which were identified among the patients who visited pediatric emergency department for nontraumatic dental conditions, include cost of treatment, lack of insurance, and limited pool of dental health care practices willing to take Medicaid-enrolled patients. 21 Even though the present study could not explore the socioeconomic status of treated patients, the literature reveals that cellulitis of the mouth occurs more in individuals of lower socioeconomic status. 22 It is likely that those who come from low socioeconomic status may have poor access to regular dental care and/or have lack of awareness in the importance of timely management of dental infections. Therefore, public education on odontogenic infection needs to be implemented and promoted, especially for those who are at increased risk for developing cellulitis of the mouth. As noted in the present study, diabetes is associated with a significant increase in both hospitalization charges and length of stay among the admitted patients. Although diabetic patients are educated on taking extra inspections to recognize any problem areas and consider adequate interventions to prevent subsequent infections, 23 there is lack of readily available resources and information on odontogenic infection and its potential for spread of the infection for diabetic patients and the general public. Despite the rare nature of systemic odontogenic infection complications, bacteremia can more readily occur in a certain cohort of patients after invasive dental procedures, such as tooth extractions and periodontal surgeries. 24 Seppanen et al. have also found that medically compromised patients are likely to undergo systemic infection complications resulting in longer hos-

7 OOOO ORIGINAL ARTICLE Volume 113, Number 1 Kim et al. 27 pital stays and an increased utilization of hospital resources. 25 Therefore, it is especially important for a dental care provider to give additional attention to compromised patients with odontogenic infection and deliver proper treatments to prevent further spread of infection that can result in hospitalization. Multiple studies have described a substantial amount of hospital resources being used for patients admitted for dental conditions. Allareddy et al. discuss the factors affecting hospitalization outcomes for patients with periapical abscesses, in addition to highlighting the total hospital charges ($105.8 million) and total hospitalization days (23,001 days) associated with admitted patients primarily due to periapical abscesses. 26 Moreover, there was a significant increase in visits and admissions for dental care in a 5-year retrospective study of pediatric emergency department visits, where more than two-thirds of the admissions were the result of untreated dental caries and subsequent abscesses and or cellulitis. 11 Such findings, along with those of the present study, indicate the trend in increased use of hospital resources for dental-related conditions. Therefore, there should be a combined effort among health care providers and policy makers in identifying and addressing areas that affect hospitalization resources. The present study is a retrospective analysis of hospitalizations on a national level; therefore, inherent limitations to studies involving secondary datasets exist. Owing to the lack of data on cellulitis of the mouth related hospitalizations in community centers, for instance, the total hospitalization charges and length of stays could be an underestimate. In addition, the dataset did not include charges associated with medications, outpatient costs, out-of-hospital costs, and other indirect costs. Therefore, the overall hospitalization resources associated with abscess or cellulitis of the mouth shown in this study should be interpreted with caution. Selection of cases included in the current study was based on the single ICD-9-CM code of 528.3, which refers to cellulitis or abscess of mouth. Additional ICD-9-CM codes, such as 682.0, which refers to cellulitis of face, were not included in the present study. The focus of our study was to examine a narrow set of conditions that could have an odontogenic origin. Including a wide range of ICD-9-CM codes would lead to heterogeneity of included cases and a possible clustering of outcomes within conditions which could be difficult to control for in our regression models. On the other hand, including only 1 condition could have resulted in an underestimation of the actual number of cases. One another major issue with the dataset is that the actual etiology of the primary diagnosis condition is not known. The dataset does not provide adequate levels of clinical information for us to determine the etiology of the condition under study. This study does not capture the causative factor for developing abscess and/or cellulitis of the mouth, which could predispose the patients at a higher risk for developing complications. Finally, the data collection process and coding practices of hospitals are likely to vary, and our results should be interpreted while keeping the above-mentioned limitations in perspective. CONCLUSIONS This study examines outcomes in patients hospitalized for cellulitis or abscess of mouth. Multiple comorbid conditions were associated with significant increase in hospitalization charges and lengths of stay among the admitted patients. Future studies should concentrate on identifying a high-risk patient population for developing spread of odontogenic infection that warrants hospitalization care. The quality of communications shared by dental care providers and patients regarding odontogenic infection management needs to be explored as well. Continuous efforts need to be made for health care providers to arrive at early diagnosis and to deliver proper antibiotic therapy and surgical drainage to eliminate the cause of infection. REFERENCES 1. Zaleckas L, Rasteniene R, Rimkuviene J, Seselgyte R. Retrospective analysis of cellulitis of the floor of the mouth. Stomatologija 2010;12: Eron L, Laine C, Goldmann D, Sox H. In the clinic: cellulitis and soft-tissue infections. Ann Intern Med 2009;150(1):ITC Fritsch DE, Klein DG. Ludwig s angina. Heart Lung 1992; 21: Huang TT, Liu TC, Chen PR, Tseng FY, Yeh TH, Chen YS. Deep neck infection: analysis of 185 cases. Head Heck 2004; 26: Zeitoun IM, Dhanarajani PJ. Cervical cellulitis and mediastinitis caused by odontogenic infections: report of two cases and review of literature. J Oral Maxillofac Surg 1995;53: Biederman GR, Dodson TB. Epidermiologic review of facial infections in hospitalized pediatric patients. J Oral Maxillofac Surg 1994;52: Boscolo-Rizzo P, Da Mosto MC. Submandibular space infection: a potentially lethal infection. Int J Infect Dis 2009;13: Marcus BJ, Kaplan J, Collins KA. A case of Ludwig angina: a case report and review of the literature. Am J Forensic Med Pathol 2008;29: Green AW, Flower EA, New NE. Mortality associated with odontogenic infection! Br Dent J 2001;190: Wills PI, Vernon RP Jr. Complications of space infections of the head and neck. Laryngoscope 1981;91: Ladrillo TE, Hobdell MH, Caviness AC. Increasing prevalence of emergency department visits for pediatric dental care, J Am Dent Assoc 2006;137: Moles DR, Ashley P. Hospital admissions for dental care in children: England Br Dent J 2009;206:E Thikkurissy S, Rawlins JT, Kumar A, Evans E, Casamassimo PS. Rapid treatment reduces hospitalization for pediatric patietns with odontogenic-based cellulitis. Am J Emerg Med 2010;28:

8 OOOO 28 Kim et al. January Kinzer S, Pfeiffer J, Becker S, Ridder GJ. Severe deep neck space infections and mediastinitis of odontogenic origin: clinical relevance and implications for diagnosis and treatment. Acta Otolaryngol 2009;129: Agency for Healthcare Research and Quality. Healthcare Cost and Utilization Project (HCUP), Nationwide Inpatient Sample (NIS), year 2008 documentation. Available at: Accessed Nov. 25, Fielding F, Reck F, Barker J. Use of magnetic resonance imaging for localization of a maxillofacial infection: report of a case. J Oral Maxillofac Surg 1987;45: Flynn TR, Shanti RM, Hayes C. Severe odontogenic infections, part 2: prospective outcomes study. J Oral Maxillofac Surg 2006;64: Kremer MJ, Blair T. Ludwig angina: forewarned is forearmed. AANA J 2006;74: Reynolds SC, Chow AW, Severe ST. Infections of the head and neck: a primer for critical care physicians. Lung 2009;187: Hwang T, Antoun JS, Lee KH. Features of odontogenic infections in hospitalized and nonhospitalised settings. Emerg Med J [Epub ahead of print; Nov 2, 2010]. 21. Dorfman DH. Dental concerns unrelated to trauma in the pediatric emergency dept: barriers to care. Arch Pediatr Adolesc Med Agarwal AK, Sethi A, Sethi D, Mrig S, Chpra S. Role of socioeconomic factors in deep neck abscess: a prospective study of 120 patients. Br J Oral Maxillofac Surg 2007; 45: The Counsel of the State Governments. Keeping people with diabetes healthy: legislator policy brief. Available at: Accessed Nov. 9, Rajasuo A, Nyfors S, Kanervo A, Jousimies-Somer H, Lindqvist C, Suuronen R. Bacteremia after plate removal and tooth extraction. Int J Oral Maxillofac Surg 2004;33: Seppanen L, Lauhio A, Lindqvist C, Suuronen R, Rautemaa R. Analysis of systemic and local odontogenic infection complications requiring hospital care. J Infect 2008;57: Allareddy V, Lin CY, Shah A, Lee MK, Nalliah R, Elangovan S, et al. Outcomes in patients hospitalized for periapical abscess in the United States: an analysis involving the use of a nationwide inpatient sample. J Am Dent Assoc 2010;141: Reprint requests: Min Kyeong Kim, BA Harvard School of Dental Medicine 188 Longwood Avenue Boston, MA minkyeong_kim@hsdm.harvard.edu