Original Contribution

Rapid treatment reduces hospitalization for pediatric patients with odontogenic-based cellulitis^?

Sarat Thikkurissy DDS, MS ^a,?, Joseph T. Rawlins DDS, MS ^b, Ashok Kumar DDS, MS ^a,

Erik Evans DDS, MD ^c, Paul S. Casamassimo DDS, MS ^a

^aDivision of Pediatric Dentistry, The Ohio State University College of Dentistry, Department of Dentistry,

Nationwide Children’s Hospital, Columbus, OH, USA

^bPrivate Practice, Rocklin, CA, USA

^cDivision of Oral and Maxillofacial Surgery, The Ohio State University College of Dentistry, Department of Dentistry, Nationwide Children’s Hospital, Columbus, OH, USA

Received 19 January 2009; accepted 26 February 2009

Abstract

Purpose: The study aimed to assess characteristics of facial cellulitis admissions and their relationship to cost of hospitalization (COH) and length of stay (LOS) in children ages 0 to 20 years at an urban hospital and to compare outcomes of rapid management to published and national statistics for LOS and COH. Methods: A retrospective review of 376 charts of facial cellulitis admissions between 2000 and 2006 revealed 63 of confirmed odontogenic cases from which cellulitis characteristics, COH, and LOS were gleaned. Variables were correlated to LOS and COH. Data on LOS and cost of admission were compared to published studies and 506 entries from the 2006 Kids’ Inpatient Database (KID).

Results: Of 63 charts included, children included were 8.3 years (SD, +-3.8 years) and equal in sex distribution. Treatment rendered and site of infection had no significant relationship to COH. Overall mean hospital LOS was 2.08 days and significantly less as compared to 3.97 days for published studies and 3.4 days for KID (P b .0001). The mean overall hospital COH was $4166 and significantly less compared to

$3223 in the literature and $8998.43 for KID.

Conclusion: In the management of pediatric facial cellulitis of odontogenic origin, rapid treatment had a significant positive impact on length of stay and Total cost of treatment compared to published studies and nationally reflective data.

(C) 2010

Introduction

Despite a reduction in permanent Dental Caries in recent years, more than half of all children have the disease by second grade, and by the time students finish high school, about 80% have caries [1]. Over time, if caries remains

^? The authors would like to acknowledge the efforts of Wei Wang and the support of the Nationwide Children’s Hospital Research Institute.

* Corresponding author.

untreated, infection may disseminate beyond the dentition to produce a dentoalveolar abscess, which spreads into surrounding connective tissue as a cellulitis, with wide clinical variability due to involvement of multiple anatomic structures, polymicrobial etiology, and differing Disease progression. Accurate diagnosis and effective treatment can be challenging and costly because of this variability but is critical in successful management as delay can spread the infection to vital anatomic structures or systemically, producing sepsis or meningitis and hospital admission [2-4].

0735-6757/$ – see front matter (C) 2010 doi:10.1016/j.ajem.2009.02.028

Facial cellulitis is classified as odontogenic or non- odontogenic based on source of infection. Odontogenic cellulitis includes infections from the dentition, and nonodontogenic cellulitis usually arises from trauma, sinus or skin infections, or idiopathic causes [5,6]. Bieder- man and Dodson [7] found that odontogenic cellulitis comprised approximately 50% of total hospital facial infections over a 10-year period.

In addition to treating serious odontogenic facial infections aggressively with parenteral antibiotics, more definitive treatment should be considered at the time of hospital admission, but often is not [8]. Sole reliance on antibiotics may be due to lack of access to dental services or an available dental service not being consulted. Pettinato et al [8] reported that no dental consult was requested in 69% of dental emergency department (ED) visits with pediatric dental residents’ available. Definitive care of odontogenic cellulitis as soon as possible after admission may decrease the total hospital length of stay and overall cost of hospitalization (COH), which can be sizable. Ettelbrick et al [1] reported that the mean hospital admission cost with children admitted to 5 different children’s hospitals in 1997 because of caries was $3223.00.

In our institution, the protocol for resolving serious facial cellulitis of odontogenic origin is stepwise, beginning with confirmation of odontogenic focus, admission, intra- venous antibiotics, and removal of the offending tooth as soon as feasible. Definitive treatment is done under local anesthesia in a clinic setting or under general anesthesia in an operating suite. Factors such as age, nature of the odontogenic infection, patient status, and social issues (ie, distance traveled) are all taken into consideration when treatment recommendations are made. In-patient manage- ment of facial cellulitis is accomplished by the infectious disease or general medicine service with surgical care provided by the dental department, and the decision when to treat is made jointly.

Previous research on odontogenic facial cellulitis has focused on the epidemiology, characteristics, and economics of the disease. Several studies have examined important variables in odontogenic facial infections such as age, sex, admission temperature, and LOS; however, none have looked at the impact that rapid treatment may have on course and cost of treatment. The aims of this study were to (1) assess characteristics and treatment approaches for facial cellulitis in an urban children’s hospital with a 24-hour dental service, (2) determine their relationship to LOS and COH, and (3) compare COH and LOS outcomes with published reports and recent national data when a rapid treatment protocol was used.

Methods

This study involved a 3-part data collection. The first was an institutional review board-approved retrospective

cohort study of pediatric patients younger than 20 years admitted for management of odontogenic facial infections during a 6-year period from 2000 to 2006. Patients were identified using distinct International Classification of Diseases, Ninth Revision (ICD-9) codes associated with infectious disease and facial infections. This set of codes can be grouped into the following Diagnostic categories: facial, orbital, or oral/perioral cellulitis (ICD code 522); inflammation of the pulpal, periapical, gingival, or period- ontal tissues (ICD code 523); jaw or salivary gland inflammation (ICD code 526,527); and facial infection (ICD code 682 ). Using these codes, computer searches were conducted for patients admitted during the study period, yielding a preliminary sample of 376. These 376 patient records were then assessed for completeness of medical information on the admission and a dental record to confirm the odontogenic nature of the cellulitis from radiographs and examination, yielding a working sample of

63 confirmed odontogenic cellulitis records. Variables abstracted from medical and dental records included age, sex, distance traveled, admission temperature, month of admission, and LOS (in days; day of admission = day 0). Odontogenic variables included tooth of origin using universal notation (nos.1-32; A-T), and dental arch (maxillary, mandibular, or both). Treatment variables analyzed during hospital admission included location of infection (body of mandible, submandibular, crosses mid- line, loss of nasolabial fold, loss of nasogenial crease), first service consulted (ED, infectious disease, pediatric den- tistry, oral/max surgery, other), additional services con- sulted, treatment performed (extraction, root canal, extraction/drain, root canal/drain, medication only). Addi- tional data included day treatment was performed compared to day of admission (treatment performed same day as admission = day 0), whether a culture was performed, imaging performed (periapical, panoramic, computed tomography, ultrasound, other), and use of general anesthesia. Antibiotic used during the hospital stay and given at the time of discharge was recorded as was the analgesic given during admission and at the time of discharge. Finally, total hospital charges (ED, facility, pharmacy, and dental) and Type of insurance were determined. Facility charges included total dollar amount for all tests ordered and charges for all nursing care required. Pharmacy charges included medications given during hospital stay. Dental charges included treatment performed on the odontogenic source of infection. Physi- cian charges were not available.

The second part of the study involved accessing the Kids’ Inpatient Database (KID). The KID is a subset of the health care utilization Project, which is sponsored by the Agency for Healthcare Research and Quality. The KID records hospital inpatient stays for children. The KID is the only all-payer inpatient care database for children in the United States. The KID contains up to 3 million hospital discharges for children ages 0 to 18 years. Data for this

Table 1 (continued)

Variable Results

Pain medication during course of 38.1% (24) admission (some patients Acetaminophen received N1 medication) 15.9% (10) Ibuprofen

15.9% (10)

Acetaminophen + ibuprofen

6.3% (4) No pain medications

4.8% (3) Acetaminophen

+ codeine

4.8% (3) Morphine

4.8% (3) Acetaminophen

+ ibuprofen + morphine 3.2% (2) Ibuprofen + morphine

3.2% (2) Acetaminophen

+ codeine + morphine 1.6% (1) Acetaminophen

+ codeine 1.6% (1) Other

Discharge pain medication 38% (10) Acetaminophen

Yes = 41.2% (26) 27% (7) Ibuprofen

No = 58.7% (37) 19% (5) Acetaminophen + codeine

8% (2) Acetaminophen + codeine + ibuprofen

4% (1) Acetaminophen + ibuprofen

4% (1) Other

study were drawn from the 2006 KID, which draws information from 38 State Inpatient Databases on children

Table 1 Variables from hospital sample Variable Results

Age (in years) 8.3 (SD, 3.8)

Sex 50.7% (32) Female

49.3% (31) Male

Distance traveled (miles) Mean, 39.4 44.4% (28) 0-10

19.0% (12) 10-20

15.8% (10) 20-50

20.6% (13) N50

Method of payment 77.6% (49) Public insurance

19.0% (12) Private insurance

3.4% (2) Self-pay

First service consulted 61.9% (39) Pediatric dentistry

38.1% (24) Emergency medicine

Imaging (some patients had N 1 80.9% (51) Periapical

image) 34.9% (22) Panoramic

19.0% (12) Computed tomography

4.8% (3) Other

1.6% (1) Ultrasound

Maxillary vs mandibular 0-71 mo: 82% maxilla

N72 mo: 55% maxilla

Location of infection 39.6% (25) Unilateral upper face

31.7% (20) Unilateral lower face

17.4% (11) Bilateral upper face

9.5% (6) Upper and lower face

1.6% (1) Bilateral lower face

Admission temperature 99.6 F (SD, 1.5)

Treatment performed (intraoral 74.6% (47) Extraction drains may have been placed in 17.4% (11) Pulpectomy conjunction with other treatment) 7.9% (5) No treatment

4.8% (3) Intraoral drain placement

Day of treatment 36.5% (23) Day 0 (day of initial presentation) 33.3% (21) Day 1

11.1% (7) Day 2

7.9% (5) No treatment

6.3% (4) Day 4 or later

4.8% (3) Day 3

Antibiotics during course of 58.7% (37) Intravenous admission Cleocin

41.2% (25) Intravenous Unasyn

Discharge antibiotics (all oral 57.1% (36) Augmentin forms) 33.3% (21) Clindamycin

3.2% (2) Penicillin VK

1.6% (1) Amoxicillin

20 years and younger. Data for ICD-9 codes 528.3 (cellulitis and abscess of oral Soft tissues) and 522.5 (periapical abscess without sinus involvement) were

collapsed for presented data.

Finally, a literature search was conducted to identify studies of odontogenic cellulitis in children. Minimal criteria for inclusion in this study were (1) treatment of children in primary, mixed, and permanent dentition through age 20 years; (2) documentation of the odontogenic origin of the cellulitis; and (3) data on COH and LOS and publication no earlier than 1980. Literature was obtained using a number of search methods including use of PubMed. The PubMed database was searched from 1980 to November 2008. Searches were limited to human studies on subjects younger

than 18 years published in English.

A database was created using Excel (Microsoft Corp, Redmond, WA) spreadsheet and then analyzed using GraphPad (GraphPad 5.1, San Diego, Calif). Bivariate statistics (t tests and ?2 statistics) were computed to assess the differences between various study variables. Differences were considered to be statistically significant if P <= .05. In addition, a multivariate logistic regression was performed comparing every variable to total cost and LOS.

Results

Hospital data

Results of variables addressed from the hospital sample of

63 subjects are presented in Table 1. The mean age of children was 8.3 years with an SD of 3 years 8 months. The overall mean LOS was 2.09 days and the mean COH was

$4166. Statistical analysis reveal there was a statistically significant difference with respect to sex and LOS (P = .048), as well as for those patients who had the tooth extracted vs receiving pulp therapy (P = .035).

Patients traveled a mean of 39.4 miles for treatment, with 22% traveling greater than 50 miles. Distance traveled was not significantly correlated to LOS (P = .46). There was no correlation between location of infection and LOS (P = .91). There was no significant correlation between LOS and type of tooth involved (P = .17), or type of tooth (permanent or primary) P = .051. There was no significant relationship between choice of antibiotic and LOS (P = .79). Definitive treatment vs just antibiotics had no significant impact on LOS (P = .35), although there was a significant difference between extraction and initiation of endodontic root canal therapy (P = .035).

Kids’ Inpatient Database data

The mean LOS noted was 3.38 days (0.184). The KID also allows one to break admissions into types of hospital, as Nationwide Children’s Hospital is an urban hospital–the urban KID data are also included; the mean LOS was 3.46 days (0.21).

Total cost

• • 1. Hospital

The overall mean total cost was $4166.00 (SD,

$2401.33). If general anesthesia was required for treatment, an additional $3083.00 was required, which significantly impacted the total cost (P = .0008). Treatment of all caries in addition to offending tooth did not significantly increase the overall total cost (P = .986). There was a significantly proportional relationship between LOS and total cost (P b .0003).

Table 2 Comparative findings of pediatric cellulitis studies from the past 20 years

• • 1. Kids’ Inpatient Database data

The overall mean total cost was $8998.43 (713). The mean for urban hospitals was $9850.70 (SD, 834.3). A breakdown of the KID total cost data is presented along with a comparison of hospital data, KID data, and the literature in Table 2.

Discussion

The purposes of this study were to assess characteristics of odontogenic cellulitis and its management and then determine whether these affected cost or LOS. The third objective was to compare the hospital findings to national statistics and published reports. The findings of this study suggest that both LOS and total cost are significantly reduced when dental and medical services manage care of these patients.

Epidemiology

Few reports describe pure odontogenic infections in the pediatric population, challenging determination of an age for which children are most likely to incur an odontogenic infection requiring hospitalization. In this study, the mean age was 8.3 +- 3.8 years, similar to Unkel et al who reported a mean age of 8.8 +- 4.4 years. Lin et al reported an average age of 5.7 +- 2.7 years in an odontogenic infection population and Ettelbrick et al a mean age of 7 years. Dodson et al found an average age of 4.55 +- 0.4 years, but this included both odontogenic and nonodontogenic infections. These varia- tions may simply reflect the differences in the populations studied, the contamination of data by nonodontogenic cellulitis, or other factors. The KID data demonstrated a rise in costs related to age; however, this may be explained through the nature of dental treatment or scope of treatment. Chow reported that orofacial odontogenic infections travel into the anatomic spaces of least resistance. Our study found most odontogenic cases were in the upper face, with 56% located purely in the upper face, and additional 10% in both upper and lower face. These findings are consistent with previous literature that similarly found that upper face infections outnumbered lower facial infections. However, contrary to previous literature that reported an

Year		Sample size	Mean age (y)	LOS (d)	Total COH
Unkel [9]	1980-1989	100	8.8	N/A	N/A
Dodson [6]	1982-1986	113	4.6	3.5	N/A
Ettelbrick [1]	1997	52	7	3.4	$3223
Lin [10]	2003	56	5.7	5.0	N/A
KID	2006	506	4	3.4	$8998.43
NCH Hospital	2000-2006	63	8.3	2.1	$4166
N/A indicates not applicable.

increased LOS and morbidity in lower face infections, we found no significant difference in the overall LOS between upper and lower facial infections (P = .914). We attribute this to our rapid treatment approach vs a wait-and-see approach, using a largely nonsurgical antibiotic-dominated pathway of care in other studies.

Management

Our standard of care is to treat the offending tooth/teeth as soon as possible with extraction or pulp therapy along with Intravenous antibiotics. Timing of treatment is based on patient medical stability, access to the oral cavity’s surgical site, and availability of surgical time if general anesthesia is needed. No significant differences in LOS were seen when the tooth was treated on the day of admission vs later in the hospital stay, although that time span was usually very short. Traditionally, the standard of care as reflected in literature is to treat cellulitis with aggressive antibiotics during the hospital stay and then treat the offending teeth after discharge, but this risks decreased follow-up to eliminate the source of infection with possible recurrence and Prolonged LOS. To our knowledge, this is the only Pediatric ARTIcle describing early treatment and elimination of source of the infection at the time of admission or as soon thereafter as possible, so it is of no surprise that the LOS for this study was 2.08 +- 1.02 days. This mean LOS is substantially lower than previous reports that range from 4.4 to 8.9 days. Furthermore, the hospital data are significantly lower than data from the 2006 KID (P b .0001). Dodson et al reported an average LOS of 3.5 days but included both odontogenic and nonodontogenic pediatric samples, and only 1 of 22 odontogenic cellulitis patients had a LOS shorter than 4 days. We propose that the dramatic decrease in the overall LOS in our study is due to the aggressive, definitive treatment given at the time of admission by an active dental service. This eliminates the primary source of the infection allowing antibiotics to theoretically work at a faster rate of resolution.

As reported previously, the mean LOS was 2.0 +- 0.99 days. The overall mean total cost to treat patients admitted of odontogenic facial cellulitis was $4138 +- 2376. This cost is higher than the previously reported $3223 by Ettelbrick et al in 2000. Inflation and rising health care costs could account for the discrepancy, but if general anesthesia was needed to treat patients because of cooperation or morbidity, a

significant cost increase of $3083 +- 1199 was required (P = .0008). Interestingly, treatment of all additional caries in addition to the offending tooth did not significantly increase the overall cost of treatment (P = .9860). We conclude that in cases when general anesthesia is required, all additional caries should be treated to improve overall health and decrease the chance for readmission.

Shortcomings in the method of analysis presented include those inherent in using data derived from multiple studies and databases retrospectively, such as assumptions made about the activity of dental services at hospitals included in the KID data and use of data skewed toward an older cohort. No data are available about postoperative complications or readmission to the hospital.

Conclusion

This study suggests that when a rapid treatment protocol is applied to pediatric patients, LOS and COH are reduced against those published in the literature and available in national databases.

References

1. Ettelbrick K, Webb M, Seale S. Hospital charges for dental caries related emergency admissions. Pediatr Dent 2000;22:1.
2. National Center for Health Statistics (NCHS). National Health and Nutrition Examination Survey III, 1988-1994. Hyattsville (Md): Centers for Disease Control and Prevention (CDC), unpublished data.
3. Jackson K, Baker SR. PeriOrbital cellulitis. Head Neck Surg 1987; 9:227.
4. Baker RC, Bausher JC. Meningitis complicating acute bacteremic facial cellulitis. Pediatr Infect Dis J 1986;5:421-3.
5. Israele V, Nelson JD. Periorbital and orbital cellulitis. Pediatr Infect Dis J 1988;6:404.
6. Dodson TB, Perrott DS, Kaban LB. Pediatric maxillofacial infections: a retrospective study of 113 patients. J Oral Maxillofac Surg 1989;47: 327-30.
7. Biederman GR, Dodson TB. Epidemiologic review of facial infections in hospitalized pediatric patients. J Oral Maxillofac Surg 1994;52: 1042-5.
8. Pettinato E, Webb M, Seale S. A comparison of Medicaid reimburse- ment for non-definitive pediatric dental treatment in the emergency room versus periodic Preventive care. Pediatric Dentistry 2000;22:6.
9. Unkel JH, Mckibben DH, Fenton SJ, et al. Comparison of odontogenic and nonodontogenic facial cellulitis in a pediatric hospital population. Pediatr Dent 1997;19:476-9.
10. Lin YT, Lu PW. Retrospective study of pediatric facial cellulitis of odontogenic origin. Pediatr Infect Dis J 2006;25:339-42.