- Email: [email protected]
Odontogenic signs and symptoms as predictors of odontogenic infection
R
E
S
E
A
R
C
H
Odontogenic signs and symptoms as predictors of odontogenic infection A clinical trial Michael T. Brennan, DDS, MHS; Michael S. Runyon, MD; Jayne J. Batts, MD; Philip C. Fox, DDS; M. Louise Kent, RN; Timothy L. Cox, DDS; H. James Norton, PhD; Peter B. Lockhart, DDS
JADA, Vol. 137
http://jada.ada.org
✷
D
A
✷
A
T
IO N
Background. The authors conducted a study to determine if odontogenic signs and symptoms in the N C emergency department predicted the development of U A ING EDU 2 overt odontogenic infection at a follow-up dental visit. RT ICLE Methods. One hundred ninety-five patients with odontalgia, but without overt signs of infection, were enrolled in a prospective, double-blind, randomized clinical trial. Data included dental diagnosis, pain characteristics, presence of caries and restorations, presence and size of periapical radiolucencies and other diagnostic test results. Results. Thirteen of 134 subjects for whom data were available had signs of infection at the follow-up visit. Subjects in the follow-up infected (FU-I) group had larger baseline radiolucencies than did subjects in the follow-up noninfected (FU-NI) group, and restorations were more prevalent for involved teeth in the FU-I group than in the FU-NI group. Conclusions. A relationship exists between radiolucency size and the presence of amalgam restorations in patients who develop clinical signs of infection. Penicillin did not appear to influence this progression. Clinical Implications. Antibiotics are not effective in preventing the development of odontogenic infection when definitive dental therapy cannot be provided for acute pain in the absence of clinical signs of infection. Although the overall risk of developing infection is low, early treatment is indicated for teeth with larger periapical radiolucencies, amalgam restorations or both. Key Words. Infection; antibiotics; toothache; risk factors. JADA 2006;137:62-6. I
62
A
J
T
P
ABSTRACT CON
atients with acute odontogenic pain frequently seek care in hospital emergency departments (EDs). Given the limitations in materials, equipment and dental staffing, definitive care rarely is delivered in this setting. This scenario is common in general dental practice as well, in that a busy patient schedule may not allow for definitive dental treatment at the initial appointment. However, the treating physician or dentist must evaluate the patient carefully for the presence of an odontogenic infection and initiate appropriate pain management. Antibiotics often are indicated for the treatment of overt odontogenic infections, but it is estimated that antibiotics are used inappropriately in 75 percent of cases involving dental conditions without signs of infection.1,2 We conducted a clinical trial to test our hypothesis that penicillin is not beneficial in the prevention of odontogenic infection in patients with acute dental pain but without clinical signs of infection. The results of that randomized, double-blind, placebo-controlled study demonstrate that antibiotics have no effect in preventing an odontogenic infection within one week.3 The purpose of the current study was to determine which presenting signs and symptoms were associated with the subsequent development of overt odontogenic infection.
Dr. Brennan is the director, Oral Medicine Residency, Carolinas Medical Center, Department of Oral Medicine, 1000 Blythe Blvd., MEB-409, Charlotte, N.C. 28232, e-mail “[email protected]”. Address reprint requests to Dr. Brennan. Dr. Runyon is the director, Medical Student Education, Emergency Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Batts is the director, Clinical Forensic Medicine, Emergency Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Fox is the director, Clinical Research, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C. Ms. Kent is the oral medicine clinical coordinator, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Cox is the director, General Practice Residency, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Norton is the director, Biostatistics, Department of Biostatistics, Carolinas Medical Center, Charlotte, N.C. Dr. Lockhart is the chairman, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C.
January 2006
Copyright ©2006 American Dental Association. All rights reserved.
R E S E A R C H
SUBJECTS AND METHODS
We conducted a randomized, double-blind, placebo-controlled trial of adult patients in a busy ED (105,000 visits per year) at an urban teaching hospital. Patients with dental pain but without overt signs of dental infection were enrolled from February 2000 to October 2001. Patients were randomized via a computer-generated numbering scheme to receive penicillin (500 milligrams) or a placebo four times per day for seven days. Exclusion criteria included the following: overt signs of infection (fever, intraoral or extraoral swelling, purulence and trismus), pregnancy, immunocompromised condition, heart disease requiring antibiotic prophylaxis, antibiotic use within one month of the study or a history of acute dental trauma. All subjects reviewed and signed a consent form approved by the institutional review board of Carolinas Medical Center, Charlotte, N.C. Baseline data. Attending or resident emergency physicians identified and screened study patients for exclusion criteria. All eligible patients underwent a structured evaluation by a dental resident at the initial ED visit. The dental resident collected the following baseline data: demographic information; pain description, intensity (measured by a visual analog scale [VAS]) and duration; dental and periodontal assessments; results of diagnostic tests, including cold, bite stick and percussion; as well as periapical radiographic findings. Patients indicated pain intensity on a VAS, with “no pain” and “worst pain” as the anchors. For pain description, the data entry sheet contained questions regarding whether the pain was stimulated versus spontaneous. If the patient reported that the pain was stimulated, the dental resident questioned him or her further regarding the following: whether the cold stimulus was fleeting versus lingering (< 10 seconds versus ≥ 10 seconds), whether the sweet stimulus was fleeting versus lingering (< 10 seconds versus ≥ 10 seconds) and whether the hot stimulus was fleeting versus lingering (< 10 seconds versus ≥ 10 seconds). The dental resident also asked patients to describe the severity and character of their pain and recorded this information on the data entry sheet. A dental resident documented caries depth and the presence and size of periapical radiolucencies. He or she calculated the size of the periapical
radiolucency as the greatest distance from the root to the border of the radiolucency. One of us (M.B.), who was a blinded investigator, confirmed the size of the radiolucency. We determined the dental diagnosis on the basis of the patient’s pain history and clinical and radiographic examination findings. Using objective and subjective data from the initial ED visit, we diagnosed patients with a symptomatic vital pulp and no signs of periapical changes as having irreversible pulpitis, and we gave a diagnosis of an abscess to patients with a nonvital pulp and periapical changes. We used a cold stimulus test (Endo-Ice, Coltène/Whaledent, Cuyahoga Falls, Ohio) in the ED to determine pulp vitality, with a positive result indicating a vital pulp and a negative result indicating a nonvital pulp. In addition to the study medication (penicillin VK versus placebo), patients were given our hospital ED’s existing standard for analgesic medications at the ED visit: 30 ibuprofen pills (800-mg tablets) and 15 hydrocodone/acetaminophen pills (5-mg/500-mg tablets). We asked subjects to return to the hospital’s dental clinic five to seven days later for the second study evaluation and definitive dental treatment. At this dental visit, a dental resident collected the following information: presence or absence of one or more of the five signs of the primary outcome variable of infection (that is, fever, intraoral or extraoral swelling, purulence and trismus); VAS pain score; medication compliance by pill count; and adverse events. Statistical analysis. We calculated descriptive statistics (including means and standard deviations) or counts and percentages. At the follow-up visit, comparisons of baseline ED data were performed between patients who had an infection (FU-I) and those with no infection (FUNI). We used a Student t test or Wilcoxon rank sum test (for ordinal data and those that were not normally distributed) to compare data measured on an interval scale between the two groups. We used the χ2 test or Fisher exact test for dichotomous data. A P value of less than .05 was considered statistically significant. We used statistical software (SAS, Version 8.2, SAS Institute, Cary, N.C.). A sample size estimate for the original, randomized, double-blind clinical trial determined that 69 subjects per group were required to detect a 15-percentage-point difference in the development of overt infection (assuming a rate of infec-
JADA, Vol. 137 http://jada.ada.org Copyright ©2006 American Dental Association. All rights reserved.
January 2006
63
R E S E A R C H
TABLE 1
Main outcome at follow-up visit.* OUTCOME
NO. (PERCENTAGE) OF SUBJECTS†
P VALUE
Placebo (n = 70)
Antibiotic (n = 64)
7 (10.0)
6 (9.4)
NS‡
0 (0)
1 (1.6)
NS
Extraoral swelling
2 (2.9)
4 (6.3)
NS
Intraoral swelling
5 (7.1)
4 (6.3)
NS
Purulence
4 (5.7)
4 (6.3)
NS
0 (0)
2 (3.1)
NS
Infection Specific Signs Fever
Trismus Mean (± SD §) Followup Pain (VAS ¶) Mild Adverse Events Mean (± SD) No. of Pills Remaining
43.5 ± 34.8 42.0 ± 31.0
NS
1 (1.4)
2 (3.1)
NS
DISCUSSION
9.3 ± 8.0
8.5 ± 7.1
NS
Odontogenic pain is a common complaint at EDs across the United States. Our ED reports approximately 4,000 such visits per year; virtually all patients are referred out for definitive treatment. These patients routinely receive analgesic and antibiotic treatment for presumed odontogenic infection or to prevent the progression to frank infection before they receive definitive dental care. Possible risk factors. A wide variety of subjective and objective clinical and radiographic variables have been proposed as risk factors for the development of postendodontic flare-up. Previous clinical trials that examined risk factors for the development of odontogenic infection were conducted in an endodontic practice setting, in which definitive endodontic treatment was provided at the first visit.4-9 In a retrospective study of 1,000 endodontic patients who had experienced an interappointment flare-up (pain, swelling or both) compared with 1,000 patients who had not experienced an interappointment flare-up, Torabinejad and colleagues4 identified the following risk factors: age, 40 to 59 years; female sex; treated mandibular first molars; presence of allergies; and preoperative pain and swelling. We found it interesting that larger periapical lesions were associated with fewer interappointment emergency visits. In another prospective study of 946 endodontic patients, Walton and Fouad5 reported that a small percentage experienced interappointment flare-ups (3.7 percent, n = 30), and they identified the following risk factors: severe pain or swelling
* Adapted with permission of Elsevier from Runyon and colleagues.3 † Unless otherwise specified. ‡ NS: Not significant. § SD: Standard deviation. ¶ VAS: Visual analog scale. 0 indicates no pain and 100 indicates worst pain.
tion of 5 and 20 percent in the penicillin and placebo groups, respectively). For subjects who did not return for the follow-up visit in the dental clinic, we searched the computerized records of our multifacility health care system for any nonstudy return visits to the ED or dental clinic. If adequate outcome data were available from these visits, we included them in the final analysis. RESULTS
We enrolled 195 subjects in the ED, with 125 (64 percent) returning for the follow-up dental care visit. For the remaining 70 subjects (36 percent) who did not return, a search of ED and dental clinic visits within one month of enrollment identified nine additional subjects who returned to our hospital owing to dental pain, infection or both. Therefore, follow-up data were available for 134 (69 percent) of 195 subjects. The main outcome—overt dental infection— was present in 13 (10 percent) of 134 subjects at the follow-up visit. As reported previously,3 we found no differences at the follow-up dental clinic visit in the outcomes of infection, pain, compliance or adverse events between patients randomized to antibiotics or to placebo (Table 1). Table 2 64
JADA, Vol. 137
presents a comparison of the baseline ED data between FU-I subjects and FU-NI subjects. There were no differences in demographic data, pain variables or dental diagnostic test results between the two groups. We found no statistical differences in the dental diagnosis at the ED visit between the FU-I and FU-NI groups. A larger percentage of patients in the FU-I group (46 percent) than in the FU-NI group (16 percent) (P = .01) had an existing amalgam restoration in the tooth identified by the clinical and radiographic ED examinations to be the source of pain and subsequent infection (Table 2). Radiographic analysis from the ED visit demonstrated that the mean size of the periapical radiolucency was larger in the FU-I group than in the FU-NI group (P = .03) (Table 2).
http://jada.ada.org
January 2006
Copyright ©2006 American Dental Association. All rights reserved.
R E S E A R C H
at the initial visit, diagnosis TABLE 2 of a necrotic pulp and an Relationship between baseline odontogenic findings acute apical abscess. and infection outcome. Neither of these studies made a distinction in the BASELINE ODONTOGENIC VARIABLE INFECTION OUTCOME P VALUE statistical analysis between No. No. a flare-up that was asso(Percentage) of (Percentage) of Subjects* With Subjects* With ciated with infection and Infection No Infection pain and a flare-up that was (n = 13) (n = 121) Mean (± SD †) Age (Years) associated with pain only, 33.4 ± 10.9 33.5 ± 10.2 NS Sex and it is not possible to NS Male 66 (55) 5 (38) – draw conclusions regarding Female 55 (45) 8 (62) – risk factors for the developRace NS ment of infection from such African-American 95 (78.5) 11 (84.6) – White 20 (16.5) 2 (15.4) small numbers of patients – Hispanic 2 (1.7) 0 – with interappointment Other 4 (3.3) 0 – flare-ups (37 [1.9 percent] of 18.9 ± 50.9 20.8 ± 27.0 Mean (± SD) Days With Pain NS 1,946). Therefore, it is diffiDiagnosis NS cult to compare these find44 (36) 8 (62) Abscess – ings with those of our study, Irreversible pulpitis 37 (31) 2 (15) – 2 (2) 0 (0) Advanced periodontal disease – which considered only infec§ 38 (31) 3 (23) Unknown – tion as the main outcome. Pain Characteristics NS Providing definitive dental 86 (71) 9 (69.2) Spontaneous – treatment is not realistic in 25 (21) 2 (15.4) Stimulated – 4 (3) 2 (15.4) Both – the ED setting. Thus, our 6 (5) 0 (0) Unknown – study reflects a more 113 (93) 12 (92) Localized Pain NS common clinical scenario and provides an opportunity Caries NS 9 (7) 0 (0) – Dentin to prospectively assess pos85 (70) 10 (77) – Pulp sible risk factors for the 12 (10) 2 (15) – None 15 (12) 1 (8) – Unknown development of an odontogenic infection. Restoration Present 19 (16) 6 (46) .01 Amalgam Use of antibiotics. The 2 (2) 1 (8) NS Resin-Based Composite literature is controversial 100 (83) 6 (46) NS None with regard to the issue of Radiolucency NS the benefit of antibiotics in 44 (36) 8 (62) Yes – 22 (18) 0 (0) No – the prevention of interap17 (14) 2 (15) Widened PDL¶ – pointment flare-ups in 38 (31) 3 (23) Unknown§ – endodontically treated, .03 1.5 ± 1.8 2.8 ± 1.8 Mean (± SD) Size of Radiolucency asymptomatic necrotic teeth (mm#) and in teeth with a localized Diagnostic Test acute apical abscess with no 70 (58) 7 (54) Cold examination, positive findings NS 11.5 ± 23.6 42.5 ± 88.7 Mean (± SD) seconds of cold NS evidence of swelling.6-10 Our NS 93 (77) 12 (92) Bite stick, positive findings data, along with those of NS 99 (82) 13 (100) Percussion, positive findings previous studies, have * Unless otherwise specified. shown no benefit from † SD: Standard deviation. antibiotics in the prevention ‡ NS: Not significant (P > .05). § The radiograph was nondiagnostic or missing. of infection before the defin¶ PDL: Periodontal ligament. # mm: Millimeters. itive dental treatment visit, thus calling into question the widespread use of antibiotics in the absence of necrotic tissue of an abscess may explain why overt infection. Fouad and colleagues10 suggested antibiotics have no effect on a localized periapical that a poor blood supply associated with the abscess. ‡
JADA, Vol. 137 http://jada.ada.org Copyright ©2006 American Dental Association. All rights reserved.
January 2006
65
R E S E A R C H
We have discussed the limitations of this prospective clinical trial elsewhere.3 Other limitations of this analysis of risk factors include the following: the low incidence of infection (n = 13), which limited the statistical assessment to a univariate analysis compared with a more ideal statistical analysis (for example, logistic regression) of baseline ED information; and our inability to determine a dental diagnosis from the objective and subjective data for 41 (31 percent) of 134 patients, owing to a missing or nondiagnostic periapical radiograph. Despite this, we had complete data for 93 patients, which allowed for statistical analysis of dental diagnostic and radiographic information, although the study was powered for the randomized, double-blind, placebo-controlled trial and not for an assessment of risk factors. Finally, we did not have a standardized protocol (for example, use of a positioning device) to ensure that the radiographic unit was used at the same angle for each periapical radiograph. This may have overestimated or underestimated the size of periapical radiolucencies. We used the same radiographic unit, processor and type of dental film in the ED for each patient. We examined whether tooth location had an impact on the presence of an amalgam restoration or the size of a periapical radiolucency. We found no differences in these two outcomes according to tooth type in the FU-I group. Six teeth with amalgam restorations became infected, including three maxillary teeth (two posterior and one anterior) and three mandibular teeth (all posterior). Our comparison of the size of periapical radiolucencies demonstrated no statistical differences between the maxillary and mandibular teeth in the FU-I group. As we reported elsewhere,3 antibiotics are neither necessary nor beneficial in the treatment of dental pain in the absence of overt infection, and eliminating this practice would reduce health care costs and minimize unnecessary antibiotic exposure. In this prospective clinical trial, few objective or subjective tests helped predict whether a patient with odontogenic pain without overt infection would develop an infection before the definitive dental treatment visit. We found it interesting that patients with an amalgam restoration in the symptomatic tooth
66
JADA, Vol. 137
http://jada.ada.org
were more likely to develop an infection before the definitive care appointment than were patients without restorations in the symptomatic tooth. This may be a spurious finding owing to the small number of subjects who developed an infection, or it may be related to prior pulpal damage resulting from removal of a large carious lesion or from preparation and placement of a prior restoration, and/or the presence of recurrent caries in an already compromised tooth. CONCLUSION
The results of this study show that patients with larger periapical radiolucencies appeared to be more likely to develop clinical signs of infection, but penicillin did not prevent progression of signs and symptoms to overt infection. Our data suggest that patients with acute dental pain eminating from a tooth with an amalgam restoration, a large periapical radiolucency or both would benefit from definitive dental treatment to prevent an infection from developing within the next five to seven days. ■ Generous financial support was provided by the Health Services Foundation Inc., Carolinas HealthCare System, Charlotte, N.C. The authors acknowledge the many general practice and oral medicine residents who treated the patients in this study and collected data for the study. 1. Dailey YM, Martin MV. Are antibiotics being used appropriately for emergency dental treatment? Br Dent J 2001;191:391-3. 2. Kandemir S, Ergül N. Grievances in cases using antibiotics due to orodental problems and assessment of the need for antibiotics. Int Dent J 2000;50(2):73-7. 3. Runyon MS, Brennan MT, Batts JJ, et al. Efficacy of penicillin for dental pain without overt infection. Acad Emerg Med 2004;11:1268-71. 4. Torabinejad M, Kettering JD, McGraw JC, Cummings RR, Dwyer TG, Tobias TS. Factors associated with endodontic interappointment emergencies of teeth with necrotic pulps. J Endod 1988;14:261-6. 5. Walton R, Fouad A. Endodontic interappointment flare-ups: a prospective study of incidence and related factors. J Endod 1992;18(4):172-7. 6. Pickenpaugh L, Reader A, Beck M, Meyers WJ, Peterson LJ. Effect of prophylactic amoxicillin on endodontic flare-up in asymptomatic, necrotic teeth. J Endod 2001;27(1):53-6. 7. Morse DR, Furst ML, Belott RM, Lefkowitz RD, Spritzer IB, Sideman BH. Infectious flare-ups and serious sequelae following endodontic treatment: a prospective randomized trial on efficacy of antibiotic prophylaxis in cases of asymptomatic pulpal-periapical lesions. Oral Surg Oral Med Oral Pathol 1987;64(1):96-109. 8. Morse DR, Furst ML, Lefkowitz RD, D’Angelo D, Esposito JV. A comparison of erythromycin and cefadroxil in the prevention of flareups from asymptomatic teeth with pulpal necrosis and associated periapical pathosis. Oral Surg Oral Med Oral Pathol 1990;69:619-30. 9. Abbott AA, Koren LZ, Morse DR, Sinai IH, Doo RS, Furst ML. A prospective randomized trial on efficacy of antibiotic prophylaxis in asymptomatic teeth with pulpal necrosis and associated periapical pathosis. Oral Surg Oral Med Oral Pathol 1988;66:722-33. 10. Fouad AF, Rivera EM, Walton RE. Penicillin as a supplement in resolving the localized acute apical abscess. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:590-5.
January 2006
Copyright ©2006 American Dental Association. All rights reserved.
E
S
E
A
R
C
H
Odontogenic signs and symptoms as predictors of odontogenic infection A clinical trial Michael T. Brennan, DDS, MHS; Michael S. Runyon, MD; Jayne J. Batts, MD; Philip C. Fox, DDS; M. Louise Kent, RN; Timothy L. Cox, DDS; H. James Norton, PhD; Peter B. Lockhart, DDS
JADA, Vol. 137
http://jada.ada.org
✷
D
A
✷
A
T
IO N
Background. The authors conducted a study to determine if odontogenic signs and symptoms in the N C emergency department predicted the development of U A ING EDU 2 overt odontogenic infection at a follow-up dental visit. RT ICLE Methods. One hundred ninety-five patients with odontalgia, but without overt signs of infection, were enrolled in a prospective, double-blind, randomized clinical trial. Data included dental diagnosis, pain characteristics, presence of caries and restorations, presence and size of periapical radiolucencies and other diagnostic test results. Results. Thirteen of 134 subjects for whom data were available had signs of infection at the follow-up visit. Subjects in the follow-up infected (FU-I) group had larger baseline radiolucencies than did subjects in the follow-up noninfected (FU-NI) group, and restorations were more prevalent for involved teeth in the FU-I group than in the FU-NI group. Conclusions. A relationship exists between radiolucency size and the presence of amalgam restorations in patients who develop clinical signs of infection. Penicillin did not appear to influence this progression. Clinical Implications. Antibiotics are not effective in preventing the development of odontogenic infection when definitive dental therapy cannot be provided for acute pain in the absence of clinical signs of infection. Although the overall risk of developing infection is low, early treatment is indicated for teeth with larger periapical radiolucencies, amalgam restorations or both. Key Words. Infection; antibiotics; toothache; risk factors. JADA 2006;137:62-6. I
62
A
J
T
P
ABSTRACT CON
atients with acute odontogenic pain frequently seek care in hospital emergency departments (EDs). Given the limitations in materials, equipment and dental staffing, definitive care rarely is delivered in this setting. This scenario is common in general dental practice as well, in that a busy patient schedule may not allow for definitive dental treatment at the initial appointment. However, the treating physician or dentist must evaluate the patient carefully for the presence of an odontogenic infection and initiate appropriate pain management. Antibiotics often are indicated for the treatment of overt odontogenic infections, but it is estimated that antibiotics are used inappropriately in 75 percent of cases involving dental conditions without signs of infection.1,2 We conducted a clinical trial to test our hypothesis that penicillin is not beneficial in the prevention of odontogenic infection in patients with acute dental pain but without clinical signs of infection. The results of that randomized, double-blind, placebo-controlled study demonstrate that antibiotics have no effect in preventing an odontogenic infection within one week.3 The purpose of the current study was to determine which presenting signs and symptoms were associated with the subsequent development of overt odontogenic infection.
Dr. Brennan is the director, Oral Medicine Residency, Carolinas Medical Center, Department of Oral Medicine, 1000 Blythe Blvd., MEB-409, Charlotte, N.C. 28232, e-mail “[email protected]”. Address reprint requests to Dr. Brennan. Dr. Runyon is the director, Medical Student Education, Emergency Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Batts is the director, Clinical Forensic Medicine, Emergency Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Fox is the director, Clinical Research, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C. Ms. Kent is the oral medicine clinical coordinator, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Cox is the director, General Practice Residency, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C. Dr. Norton is the director, Biostatistics, Department of Biostatistics, Carolinas Medical Center, Charlotte, N.C. Dr. Lockhart is the chairman, Department of Oral Medicine, Carolinas Medical Center, Charlotte, N.C.
January 2006
Copyright ©2006 American Dental Association. All rights reserved.
R E S E A R C H
SUBJECTS AND METHODS
We conducted a randomized, double-blind, placebo-controlled trial of adult patients in a busy ED (105,000 visits per year) at an urban teaching hospital. Patients with dental pain but without overt signs of dental infection were enrolled from February 2000 to October 2001. Patients were randomized via a computer-generated numbering scheme to receive penicillin (500 milligrams) or a placebo four times per day for seven days. Exclusion criteria included the following: overt signs of infection (fever, intraoral or extraoral swelling, purulence and trismus), pregnancy, immunocompromised condition, heart disease requiring antibiotic prophylaxis, antibiotic use within one month of the study or a history of acute dental trauma. All subjects reviewed and signed a consent form approved by the institutional review board of Carolinas Medical Center, Charlotte, N.C. Baseline data. Attending or resident emergency physicians identified and screened study patients for exclusion criteria. All eligible patients underwent a structured evaluation by a dental resident at the initial ED visit. The dental resident collected the following baseline data: demographic information; pain description, intensity (measured by a visual analog scale [VAS]) and duration; dental and periodontal assessments; results of diagnostic tests, including cold, bite stick and percussion; as well as periapical radiographic findings. Patients indicated pain intensity on a VAS, with “no pain” and “worst pain” as the anchors. For pain description, the data entry sheet contained questions regarding whether the pain was stimulated versus spontaneous. If the patient reported that the pain was stimulated, the dental resident questioned him or her further regarding the following: whether the cold stimulus was fleeting versus lingering (< 10 seconds versus ≥ 10 seconds), whether the sweet stimulus was fleeting versus lingering (< 10 seconds versus ≥ 10 seconds) and whether the hot stimulus was fleeting versus lingering (< 10 seconds versus ≥ 10 seconds). The dental resident also asked patients to describe the severity and character of their pain and recorded this information on the data entry sheet. A dental resident documented caries depth and the presence and size of periapical radiolucencies. He or she calculated the size of the periapical
radiolucency as the greatest distance from the root to the border of the radiolucency. One of us (M.B.), who was a blinded investigator, confirmed the size of the radiolucency. We determined the dental diagnosis on the basis of the patient’s pain history and clinical and radiographic examination findings. Using objective and subjective data from the initial ED visit, we diagnosed patients with a symptomatic vital pulp and no signs of periapical changes as having irreversible pulpitis, and we gave a diagnosis of an abscess to patients with a nonvital pulp and periapical changes. We used a cold stimulus test (Endo-Ice, Coltène/Whaledent, Cuyahoga Falls, Ohio) in the ED to determine pulp vitality, with a positive result indicating a vital pulp and a negative result indicating a nonvital pulp. In addition to the study medication (penicillin VK versus placebo), patients were given our hospital ED’s existing standard for analgesic medications at the ED visit: 30 ibuprofen pills (800-mg tablets) and 15 hydrocodone/acetaminophen pills (5-mg/500-mg tablets). We asked subjects to return to the hospital’s dental clinic five to seven days later for the second study evaluation and definitive dental treatment. At this dental visit, a dental resident collected the following information: presence or absence of one or more of the five signs of the primary outcome variable of infection (that is, fever, intraoral or extraoral swelling, purulence and trismus); VAS pain score; medication compliance by pill count; and adverse events. Statistical analysis. We calculated descriptive statistics (including means and standard deviations) or counts and percentages. At the follow-up visit, comparisons of baseline ED data were performed between patients who had an infection (FU-I) and those with no infection (FUNI). We used a Student t test or Wilcoxon rank sum test (for ordinal data and those that were not normally distributed) to compare data measured on an interval scale between the two groups. We used the χ2 test or Fisher exact test for dichotomous data. A P value of less than .05 was considered statistically significant. We used statistical software (SAS, Version 8.2, SAS Institute, Cary, N.C.). A sample size estimate for the original, randomized, double-blind clinical trial determined that 69 subjects per group were required to detect a 15-percentage-point difference in the development of overt infection (assuming a rate of infec-
JADA, Vol. 137 http://jada.ada.org Copyright ©2006 American Dental Association. All rights reserved.
January 2006
63
R E S E A R C H
TABLE 1
Main outcome at follow-up visit.* OUTCOME
NO. (PERCENTAGE) OF SUBJECTS†
P VALUE
Placebo (n = 70)
Antibiotic (n = 64)
7 (10.0)
6 (9.4)
NS‡
0 (0)
1 (1.6)
NS
Extraoral swelling
2 (2.9)
4 (6.3)
NS
Intraoral swelling
5 (7.1)
4 (6.3)
NS
Purulence
4 (5.7)
4 (6.3)
NS
0 (0)
2 (3.1)
NS
Infection Specific Signs Fever
Trismus Mean (± SD §) Followup Pain (VAS ¶) Mild Adverse Events Mean (± SD) No. of Pills Remaining
43.5 ± 34.8 42.0 ± 31.0
NS
1 (1.4)
2 (3.1)
NS
DISCUSSION
9.3 ± 8.0
8.5 ± 7.1
NS
Odontogenic pain is a common complaint at EDs across the United States. Our ED reports approximately 4,000 such visits per year; virtually all patients are referred out for definitive treatment. These patients routinely receive analgesic and antibiotic treatment for presumed odontogenic infection or to prevent the progression to frank infection before they receive definitive dental care. Possible risk factors. A wide variety of subjective and objective clinical and radiographic variables have been proposed as risk factors for the development of postendodontic flare-up. Previous clinical trials that examined risk factors for the development of odontogenic infection were conducted in an endodontic practice setting, in which definitive endodontic treatment was provided at the first visit.4-9 In a retrospective study of 1,000 endodontic patients who had experienced an interappointment flare-up (pain, swelling or both) compared with 1,000 patients who had not experienced an interappointment flare-up, Torabinejad and colleagues4 identified the following risk factors: age, 40 to 59 years; female sex; treated mandibular first molars; presence of allergies; and preoperative pain and swelling. We found it interesting that larger periapical lesions were associated with fewer interappointment emergency visits. In another prospective study of 946 endodontic patients, Walton and Fouad5 reported that a small percentage experienced interappointment flare-ups (3.7 percent, n = 30), and they identified the following risk factors: severe pain or swelling
* Adapted with permission of Elsevier from Runyon and colleagues.3 † Unless otherwise specified. ‡ NS: Not significant. § SD: Standard deviation. ¶ VAS: Visual analog scale. 0 indicates no pain and 100 indicates worst pain.
tion of 5 and 20 percent in the penicillin and placebo groups, respectively). For subjects who did not return for the follow-up visit in the dental clinic, we searched the computerized records of our multifacility health care system for any nonstudy return visits to the ED or dental clinic. If adequate outcome data were available from these visits, we included them in the final analysis. RESULTS
We enrolled 195 subjects in the ED, with 125 (64 percent) returning for the follow-up dental care visit. For the remaining 70 subjects (36 percent) who did not return, a search of ED and dental clinic visits within one month of enrollment identified nine additional subjects who returned to our hospital owing to dental pain, infection or both. Therefore, follow-up data were available for 134 (69 percent) of 195 subjects. The main outcome—overt dental infection— was present in 13 (10 percent) of 134 subjects at the follow-up visit. As reported previously,3 we found no differences at the follow-up dental clinic visit in the outcomes of infection, pain, compliance or adverse events between patients randomized to antibiotics or to placebo (Table 1). Table 2 64
JADA, Vol. 137
presents a comparison of the baseline ED data between FU-I subjects and FU-NI subjects. There were no differences in demographic data, pain variables or dental diagnostic test results between the two groups. We found no statistical differences in the dental diagnosis at the ED visit between the FU-I and FU-NI groups. A larger percentage of patients in the FU-I group (46 percent) than in the FU-NI group (16 percent) (P = .01) had an existing amalgam restoration in the tooth identified by the clinical and radiographic ED examinations to be the source of pain and subsequent infection (Table 2). Radiographic analysis from the ED visit demonstrated that the mean size of the periapical radiolucency was larger in the FU-I group than in the FU-NI group (P = .03) (Table 2).
http://jada.ada.org
January 2006
Copyright ©2006 American Dental Association. All rights reserved.
R E S E A R C H
at the initial visit, diagnosis TABLE 2 of a necrotic pulp and an Relationship between baseline odontogenic findings acute apical abscess. and infection outcome. Neither of these studies made a distinction in the BASELINE ODONTOGENIC VARIABLE INFECTION OUTCOME P VALUE statistical analysis between No. No. a flare-up that was asso(Percentage) of (Percentage) of Subjects* With Subjects* With ciated with infection and Infection No Infection pain and a flare-up that was (n = 13) (n = 121) Mean (± SD †) Age (Years) associated with pain only, 33.4 ± 10.9 33.5 ± 10.2 NS Sex and it is not possible to NS Male 66 (55) 5 (38) – draw conclusions regarding Female 55 (45) 8 (62) – risk factors for the developRace NS ment of infection from such African-American 95 (78.5) 11 (84.6) – White 20 (16.5) 2 (15.4) small numbers of patients – Hispanic 2 (1.7) 0 – with interappointment Other 4 (3.3) 0 – flare-ups (37 [1.9 percent] of 18.9 ± 50.9 20.8 ± 27.0 Mean (± SD) Days With Pain NS 1,946). Therefore, it is diffiDiagnosis NS cult to compare these find44 (36) 8 (62) Abscess – ings with those of our study, Irreversible pulpitis 37 (31) 2 (15) – 2 (2) 0 (0) Advanced periodontal disease – which considered only infec§ 38 (31) 3 (23) Unknown – tion as the main outcome. Pain Characteristics NS Providing definitive dental 86 (71) 9 (69.2) Spontaneous – treatment is not realistic in 25 (21) 2 (15.4) Stimulated – 4 (3) 2 (15.4) Both – the ED setting. Thus, our 6 (5) 0 (0) Unknown – study reflects a more 113 (93) 12 (92) Localized Pain NS common clinical scenario and provides an opportunity Caries NS 9 (7) 0 (0) – Dentin to prospectively assess pos85 (70) 10 (77) – Pulp sible risk factors for the 12 (10) 2 (15) – None 15 (12) 1 (8) – Unknown development of an odontogenic infection. Restoration Present 19 (16) 6 (46) .01 Amalgam Use of antibiotics. The 2 (2) 1 (8) NS Resin-Based Composite literature is controversial 100 (83) 6 (46) NS None with regard to the issue of Radiolucency NS the benefit of antibiotics in 44 (36) 8 (62) Yes – 22 (18) 0 (0) No – the prevention of interap17 (14) 2 (15) Widened PDL¶ – pointment flare-ups in 38 (31) 3 (23) Unknown§ – endodontically treated, .03 1.5 ± 1.8 2.8 ± 1.8 Mean (± SD) Size of Radiolucency asymptomatic necrotic teeth (mm#) and in teeth with a localized Diagnostic Test acute apical abscess with no 70 (58) 7 (54) Cold examination, positive findings NS 11.5 ± 23.6 42.5 ± 88.7 Mean (± SD) seconds of cold NS evidence of swelling.6-10 Our NS 93 (77) 12 (92) Bite stick, positive findings data, along with those of NS 99 (82) 13 (100) Percussion, positive findings previous studies, have * Unless otherwise specified. shown no benefit from † SD: Standard deviation. antibiotics in the prevention ‡ NS: Not significant (P > .05). § The radiograph was nondiagnostic or missing. of infection before the defin¶ PDL: Periodontal ligament. # mm: Millimeters. itive dental treatment visit, thus calling into question the widespread use of antibiotics in the absence of necrotic tissue of an abscess may explain why overt infection. Fouad and colleagues10 suggested antibiotics have no effect on a localized periapical that a poor blood supply associated with the abscess. ‡
JADA, Vol. 137 http://jada.ada.org Copyright ©2006 American Dental Association. All rights reserved.
January 2006
65
R E S E A R C H
We have discussed the limitations of this prospective clinical trial elsewhere.3 Other limitations of this analysis of risk factors include the following: the low incidence of infection (n = 13), which limited the statistical assessment to a univariate analysis compared with a more ideal statistical analysis (for example, logistic regression) of baseline ED information; and our inability to determine a dental diagnosis from the objective and subjective data for 41 (31 percent) of 134 patients, owing to a missing or nondiagnostic periapical radiograph. Despite this, we had complete data for 93 patients, which allowed for statistical analysis of dental diagnostic and radiographic information, although the study was powered for the randomized, double-blind, placebo-controlled trial and not for an assessment of risk factors. Finally, we did not have a standardized protocol (for example, use of a positioning device) to ensure that the radiographic unit was used at the same angle for each periapical radiograph. This may have overestimated or underestimated the size of periapical radiolucencies. We used the same radiographic unit, processor and type of dental film in the ED for each patient. We examined whether tooth location had an impact on the presence of an amalgam restoration or the size of a periapical radiolucency. We found no differences in these two outcomes according to tooth type in the FU-I group. Six teeth with amalgam restorations became infected, including three maxillary teeth (two posterior and one anterior) and three mandibular teeth (all posterior). Our comparison of the size of periapical radiolucencies demonstrated no statistical differences between the maxillary and mandibular teeth in the FU-I group. As we reported elsewhere,3 antibiotics are neither necessary nor beneficial in the treatment of dental pain in the absence of overt infection, and eliminating this practice would reduce health care costs and minimize unnecessary antibiotic exposure. In this prospective clinical trial, few objective or subjective tests helped predict whether a patient with odontogenic pain without overt infection would develop an infection before the definitive dental treatment visit. We found it interesting that patients with an amalgam restoration in the symptomatic tooth
66
JADA, Vol. 137
http://jada.ada.org
were more likely to develop an infection before the definitive care appointment than were patients without restorations in the symptomatic tooth. This may be a spurious finding owing to the small number of subjects who developed an infection, or it may be related to prior pulpal damage resulting from removal of a large carious lesion or from preparation and placement of a prior restoration, and/or the presence of recurrent caries in an already compromised tooth. CONCLUSION
The results of this study show that patients with larger periapical radiolucencies appeared to be more likely to develop clinical signs of infection, but penicillin did not prevent progression of signs and symptoms to overt infection. Our data suggest that patients with acute dental pain eminating from a tooth with an amalgam restoration, a large periapical radiolucency or both would benefit from definitive dental treatment to prevent an infection from developing within the next five to seven days. ■ Generous financial support was provided by the Health Services Foundation Inc., Carolinas HealthCare System, Charlotte, N.C. The authors acknowledge the many general practice and oral medicine residents who treated the patients in this study and collected data for the study. 1. Dailey YM, Martin MV. Are antibiotics being used appropriately for emergency dental treatment? Br Dent J 2001;191:391-3. 2. Kandemir S, Ergül N. Grievances in cases using antibiotics due to orodental problems and assessment of the need for antibiotics. Int Dent J 2000;50(2):73-7. 3. Runyon MS, Brennan MT, Batts JJ, et al. Efficacy of penicillin for dental pain without overt infection. Acad Emerg Med 2004;11:1268-71. 4. Torabinejad M, Kettering JD, McGraw JC, Cummings RR, Dwyer TG, Tobias TS. Factors associated with endodontic interappointment emergencies of teeth with necrotic pulps. J Endod 1988;14:261-6. 5. Walton R, Fouad A. Endodontic interappointment flare-ups: a prospective study of incidence and related factors. J Endod 1992;18(4):172-7. 6. Pickenpaugh L, Reader A, Beck M, Meyers WJ, Peterson LJ. Effect of prophylactic amoxicillin on endodontic flare-up in asymptomatic, necrotic teeth. J Endod 2001;27(1):53-6. 7. Morse DR, Furst ML, Belott RM, Lefkowitz RD, Spritzer IB, Sideman BH. Infectious flare-ups and serious sequelae following endodontic treatment: a prospective randomized trial on efficacy of antibiotic prophylaxis in cases of asymptomatic pulpal-periapical lesions. Oral Surg Oral Med Oral Pathol 1987;64(1):96-109. 8. Morse DR, Furst ML, Lefkowitz RD, D’Angelo D, Esposito JV. A comparison of erythromycin and cefadroxil in the prevention of flareups from asymptomatic teeth with pulpal necrosis and associated periapical pathosis. Oral Surg Oral Med Oral Pathol 1990;69:619-30. 9. Abbott AA, Koren LZ, Morse DR, Sinai IH, Doo RS, Furst ML. A prospective randomized trial on efficacy of antibiotic prophylaxis in asymptomatic teeth with pulpal necrosis and associated periapical pathosis. Oral Surg Oral Med Oral Pathol 1988;66:722-33. 10. Fouad AF, Rivera EM, Walton RE. Penicillin as a supplement in resolving the localized acute apical abscess. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1996;81:590-5.
January 2006
Copyright ©2006 American Dental Association. All rights reserved.