Clinical Research

An Investigation Into Differential Diagnosis of Pulp

and Periapical Pain: A PennEndo Database Study
Mian Iqbal, DMD, MS,* Sara Kim, DMD†, and Frank Yoon, MA‡

Abstract
Nine hundred fifty-one emergency and 997 nonemer-
gency patients seeking endodontic treatment were the
basis of this study. Variables of interest were 10 pain
D iagnostic accuracy, which refers to the ability of a test to discriminate between
subjects with and without disease, is commonly measured by using sensitivity and
specificity. When a standard diagnostic test can be used as a gold standard, determining
descriptors, percussion and palpation tests, causative the accuracy rates of other tests can be easily performed. Unfortunately, in pulpal
factors, and paired pulpal and periapical diagnoses. A diagnosis, a standard is not available to determine the accuracy of other tests. This is
higher number of patients suffering from symptomatic because no consistent relationship has been found between clinical symptoms and
pulpal conditions sought emergency care. Odds of car- histopathologic conditions of the pulp (1– 4). Patients experiencing identical clinical
ies being a causative factor were high in symptomatic symptoms may exhibit different histologic appearances of the involved pulp. The only
pulps compared with asymptomatic pulpal and peria- condition that can be diagnosed with certainty is pulp necrosis. Therefore, the clinician
pical conditions. Higher odds ratios were obtained for is forced to make a clinical diagnosis by using the patient’s history and description of
sharp pain in symptomatic pulps versus symptomatic pain, clinical examination, and various diagnostic tests.
periapical conditions. Conversely, odds ratios for dull Traditionally, descriptors of pain and other signs and symptoms have been em-
pain were higher in symptomatic periapical conditions pirically used as part of pulpal and periapical diagnosis (5– 8). A few studies have
compared with asymptomatic periapical conditions. investigated the association of signs and symptoms with clinical pulpal and periapical
Percussion and palpation tests were significant in dif- diagnoses (3, 9). However, these studies either did not use advanced statistical analyses
ferentially diagnosing between pulpal and periapical or studied small patient populations. Clearer insight into these associations might help
conditions. In conclusion, caries was associated with us discover which signs and symptoms have a potential for improving differential diag-
painful pulpitis. The results confirm the differential nosis and treatment quality in pulpal and periapical diseases.
diagnostic power of sharp and dull pain and percussion The purpose of this study is to use a logistic regression model to determine any
and palpation tests. Several symptoms previously be- associations between signs and symptoms and different pulpal and periapical diagnoses
lieved to have differential diagnostic power were found in a large patient population.
insignificant. (J Endod 2007;33:548 –551)

Keywords: Materials and Methods

Acute apical abscess, acute apical periodontitis, den- From 2000 to 2004, endodontic residents at the University of Pennsylvania re-
tinal pain, differential diagnosis, irreversible pulpitis, corded their endodontic case diagnoses in an electronic database consisting of 4,853
pain descriptors, pulpal and periapical diagnosis patients. This database contains, along with periapical and pulpal diagnoses, informa-
tion regarding patients’ pain symptoms at the time of diagnosis (binary variables for
intermittent, provoked, throbbing, lingering, spontaneous, radiating, dull, sharp, local-
ized, and diffused pain), results of clinical evaluations (palpation and percussion),
From the *Postdoctoral Endodontic Program, University of primary location of pain (tooth number), causative factors, and an indicator variable
Pennsylvania, Philadelphia, Pennsylvania; †University of Penn- for whether the case was an emergency or not. Logistic regression models were fit to
sylvania, School of Dental Medicine, Philadelphia, Pennsylva-
nia; and ‡University of Pennsylvania, Wharton College, De- these data by using the binary pain symptoms as outcomes and paired diagnoses as the
partment of Statistics, Philadelphia, Pennsylvania. primary predictor variable, controlling for emergency cases and tooth type (anterior
Address requests for reprints to Dr. Mian K. Iqbal, BDS, and posterior teeth). From these models, odds ratios for presence of each specified
DMD, MS, Postdoctoral Endodontic Program, Department of pain response at all possible pairs of diagnosis levels were generated. The significance
Endodontics, University of Pennsylvania, School of Dental
Medicine, Robert Schattner Center, 240 South 40th Street,
of these odds ratios was determined by asymptotic theory; confidence intervals were
Philadelphia, PA 19104-6030. E-mail: miqal@pobox. calculated at 95% levels, corresponding to 0.05 significance levels. Regression models
upenn.edu and all analyses performed in this study were done in the statistical package R (http:/
0099-2399/$0 - see front matter www.r-project.org).
Copyright © 2007 by the American Association of From the full database, records used for the study were selected on the basis of
Endodontists.
doi:10.1016/j.joen.2007.01.006 whether the patient indicated pain at the time of treatment. Within this subset of 2,439
records, there were 31 levels of paired diagnoses including those with missing values
either at the pulpal or periapical levels or both. Those entries with missing diagnoses
were removed for estimation purposes. In addition, it was seen that several diagnoses
achieved low counts (⬍5 cases) when the data were tabulated, and these diagnoses
were removed. The data cleanup resulted in 13 levels of paired diagnoses, with 1,948
observations, 951 emergency and 997 nonemergency cases.
The 13 levels of diagnoses resulted in 13 levels of the main predictor variable in
consideration (Table 1). This meant that odds ratios generated as described earlier

548 Iqbal et al. JOE — Volume 33, Number 5, May 2007

 Clinical Research
TABLE 1. Diagnostic Groups for Pulpal and Periapical Diagnosis
Pulpal diagnosis
Symptomatic pulpal Symptomatic irreversible pulpitis
(SPU)
Asymptomatic pulpal Asymptomatic irreversible pulpitis
(APU) Previous root canal treatment
Pulp necrosis
Periapical diagnosis
Symptomatic Primary acute apical abscess
periapical (SPE) Secondary acute apical abscess
Acute apical periodontitis
Asymptomatic Within normal limits*
periapical (APE) Chronic apical periodontitis
*Absence of swelling and normal response to percussion and palpation tests.

would produce 13 ⫻ 13 tables for each of the 10 pain characters that

would not have been easily interpretable. For purposes of practical
interpretation as well as more robust estimation, these diagnoses were
collapsed into “symptomatic” and “asymptomatic” diagnoses; thus,
from 13 levels of the main predictor variable, it was possible to reduce
the dimension to the following four levels. Figure 1. Distribution of diagnoses in combined emergency and nonemergency
patients.
1. Symptomatic diagnosis at both the pulpal and periapical levels
(SPU.SPE/symptomatic pulp and periapical). It is essentially a
combination of acute pulpitis and acute apical periodontitis. In the combined data, there were 38.5% symptomatic pulpal and
2. Asymptomatic pulpal diagnosis with an asymptomatic periapical periapical, 25.8% symptomatic periapical, 24.9% symptomatic pulpal,
diagnosis (APU.APE/or asymptomatic pulp and periapical). Its and 10.7% asymptomatic pulpal and periapical (Fig. 1). When the dis-
equivalents can range from an operative carious exposure to tribution was analyzed separately, a slightly higher percentage of asymp-
chronic apical periodontitis (Table 1). tomatic pulpal and periapical and symptomatic periapical was found in
3. Symptomatic pulpal with asymptomatic periapical (SPU.APE/ nonemergency cases, whereas the percentage of symptomatic pulpal
symptomatic pulpal). It can be considered equivalent to acute and periapical and symptomatic pulpal was found to be higher in emer-
pulpitis or irreversible pulpitis. gency patients (Fig. 2).
4. Asymptomatic pulpal with symptomatic periapical (APU.SPE/ Ten different pain characters were used to determine any signifi-
symptomatic periapical). It is basically equivalent to acute apical cant association with the four types of endodontic diagnoses. Statisti-
periodontitis or acute apical abscess. cally significant results were only obtained with sharp, dull, and linger-
ing pain (Table 3). The odds of having observed a sharp pain when the
Teeth were diagnosed as “symptomatic pulps” when they exhibited patient has been diagnosed with symptomatic pulpal and periapical and
spontaneous pain or episodes of pain to temperature changes that out- symptomatic pulpal are 3.3 and 3.8 times greater, respectively, than for
lasted the stimulus. Similarly, teeth diagnosed as “symptomatic peria-
pical” exhibited painful response to percussion and/or palpation tests
or intraoral/ extraoral swelling.

Results
Among emergency and nonemergency patients, 60.0% of root ca-
nal treatments were the result of caries, 18.4% faulty restorations,
12.9% posttreatment apical periodontitis, 3.1% trauma, 3.1% fracture,
and 1.0% idiopathic. Many patients had multiple etiologies recorded.
Because caries was the main cause for root canal treatment, further
analysis was performed to determine its relationship to diagnoses. The
odds of caries being the etiologic factor when the patient was diagnosed
with symptomatic pulpal or symptomatic pulpal and periapical was
about 4 times greater than for a similar patient diagnosed with asymp-
tomatic pulp and periapical or symptomatic periapical (p ⬍ 0.05)
(Table 2).

TABLE 2. Odds Ratios for the Presence of Caries Between Diagnoses

APU.APE SPU.APE APU.SPE SPU.SPE
Caries
APU.APE 1.000 0.245* 1.062 0.271*
SPU.APE 4.086* 1.000 4.341* 1.106
APU.SPE 0.941 0.230* 1.000 0.255*
SPU.SPE 3.693* 0.904 3.923* 1.000
Figure 2. A comparison of diagnoses between emergency and nonemergency
*Significant at the 0.05 level. patients

JOE — Volume 33, Number 5, May 2007 Pulp and Periapical Pain 549
Clinical Research
TABLE 3. Odds Ratios for Pain Descriptors Between Diagnoses numbers of diagnostic categories were used, only a small number of
APU.APE SPU.APE APU.SPE SPU.SPE variables were able to achieve statistical significance. This was similar to
previous investigators who found a poor correlation between clinical
Dull
APU.APE 1.000 3.608* 1.067 2.872* symptoms and histopathologic features when pulpal diagnosis was sub-
SPU.APE 0.277* 1.000 0.296* 0.796 divided (4, 14). Because some of the diagnostic categories used mea-
APU.SPE 0.937 3.380* 1.000 2.691* sured similar aspects, the number of diagnosis was reduced to a man-
SPU.SPE 0.348* 1.256 0.372* 1.000 ageable size.
Sharp
APU.APE 1.000 0.218* 0.840 0.252*
The results of this study indicate that 60% of root canal treatments
SPU.APE 4.577* 1.000 3.845* 1.154 were caused by carious pulp exposures. On the other hand, only 18.5%
APU.SPE 1.190 0.260* 1.000 0.300* of root canal treatments were necessitated because of restoration-re-
SPU.SPE 3.965* 0.866 3.331* 1.000 lated reasons. These results are in conformity with De Quadros et al.
Lingering
APU.APE 1.000 0.575* 0.896 1.084
(15) and Serene and Spolsky (16).
SPU.APE 1.740* 1.000 1.559 1.887 It is also interesting to note that the odds of caries being the caus-
APU.SPE 1.116 0.641 1.000 1.210 ative factor when the patient is diagnosed with symptomatic pulpal con-
SPU.SPE 0.922 0.530 0.826 1.000 ditions is about four times greater than for a similar patient diagnosed
*Significant at the 0.05 level. with asymptomatic pulpal conditions. These results are supported by
Bender (17) who reported a 79% incidence of carious pulp exposures
in the presence of severe pulpal pain. This not only points toward a
a similar patient diagnosed with symptomatic periapical (p ⬍ 0.05).
bacterial etiology of symptomatic pulpal conditions but also to the pos-
The odds of having observed a dull pain when the patient is diagnosed
sibility that in differential diagnosis a tooth with a carious lesion is
with symptomatic periapical were about 3.4 and 2.7 times greater than
responsible for symptomatic pulpal conditions.
for a similar patient diagnosed with symptomatic pulpal and periapical
or symptomatic pulpal, respectively (p ⬍ 0.05). Other pain descriptors The most common reason for seeking root canal treatment in this
such as intermittent, provoked, radiating, spontaneous, throbbing, lo- cohort of patients is symptomatic pulpal and periapical diagnosis, fol-
calized, and diffused did not show any significance with endodontic lowed by symptomatic pulpal with asymptomatic periapical diagnosis
diagnoses. (Fig. 1). The results partly conform to Saad and Clem (18) who re-
Additional results were produced for percussion and palpation ported that necrotic pulp and irreversible pulpitis were the most fre-
tests to determine any association with final diagnoses (Tables 4 and 5). quent reasons for endodontic therapy in 382 patients studied at a post-
The odds ratios were based on the event of observing a more severe graduate endodontic program. However, in our sample, almost 50% of
response in palpation and percussion tests; note that it was the direction the patients were derived from the emergency clinic.
of response that was of interest here, rather than the event of observing A comparison of emergency and nonemergency patients also in-
a discrete response, as before. For a patient with symptomatic pulpal dicates that a relatively higher percentage of patients with symptomatic
and periapical, the odds of more severe palpation and percussion re- pulpal involvement seek emergency care (Fig. 2). A small number of
sponses were 6.4 and 29.4 times greater than the same odds for a patients with asymptomatic pulpal and periapical diagnosis also sought
similar patient with symptomatic pulpal, respectively (p ⬍ 0.05). For a emergency treatment for reasons such as carious pulp exposure during
patient with symptomatic periapical, the odds of more severe palpation operative procedures.
and percussion responses were 19.9 and 48.0 times greater than for a A look at the odds ratios generated for sharp pain indicates that
similar patient with symptomatic pulpal, respectively (p ⬍ 0.05). sharp pain is primarily of pulpal origin (Table 3). One is four times
more likely to have sharp pain from a symptomatic pulpal rather than
Discussion symptomatic periapical condition. Statistically significant results involv-
Pulpal and periapical diseases constitute important groups of con- ing sharp pain were obtained only in the presence of symptomatic
ditions, but there is no universal agreement about the delineation and pulpal conditions. On the other hand, dull pain followed an opposite
classification of these disorders (10 –13). In this study, when large pattern indicating that dull pain more commonly originates from peri-
apical conditions. Laboratory and clinical findings indicate that sharp
pain is primarily mediated by A-fibers, located in the dentine, whereas
TABLE 4. Odds Ratios for Palpation Response
activation of C-fibers, located in pulp, contribute to the dull pain in-
APU. APE SPU. APE APU. SPE SPU. SPE duced by pulpal inflammation (17, 19). In this study, the odds ratios for
Palpation dull pain were 3 times greater in symptomatic periapical conditions
APU. APE 1.000 5.711* 0.287* 0.892 compared with symptomatic pulpal condition, which indicates that dull
SPU. APE 0.175* 1.000 0.050* 0.156* pain is primarily periapical in origin.
APU. SPE 3.482* 19.884* 1.000 3.105
SPU. SPE 1.121 6.403* 0.322 1.000 The results for pain provoked by cold stimuli indicated the odds of
having observed a lingering pain when the patient is diagnosed with
*Significant at the 0.05 level.
asymptomatic pulpal and periapical diagnosis are 0.6 times less than for
a similar patient diagnosed with symptomatic pulpal or asymptomatic
TABLE 5. Odds Ratios for Percussion Response periapical diagnosis (p ⬍ 0.05) (Table 3). These results reconfirm the
APU. APE SPU. APE APU. SPE SPU. SPE usefulness of thermal tests in diagnosing symptomatic pulps.
Percussion
Percussion and palpation tests were found extremely powerful in
APU. APE 1.000 10.572* 0.220* 0.359* discriminating between pulpal and periapical conditions. The odds ra-
SPU. APE 0.095* 1.000 0.021* 0.034* tios for percussion and palpation were high in symptomatic periapical
APU. SPE 4.539* 47.985* 1.000 1.631 conditions when compared with symptomatic pulpal conditions. The
SPU. SPE 2.783* 29.426* 0.613 1.000 results also show that the odds ratios for percussion and palpation are
*Significant at the 0.05 level. considerably reduced in the presence of pulpal vitality (Tables 4 and 5).

550 Iqbal et al. JOE — Volume 33, Number 5, May 2007

 Clinical Research
It is interesting to note that Klausen et al. (9), who with 92 patients pathology. Percussion and palpation tests were powerful in differentially
classified pulpal and periapical diagnoses quite comparable to ours, diagnosing between pulpal and periapical conditions. Under the condi-
were not able to distinguish between pulpoperiodontitis (symptomatic tions of this study, several symptoms previously believed to have differ-
pulpal and periapical) and pulpitis (symptomatic pulpal) on the basis of ential diagnostic power were found insignificant.
symptoms and signs included in their investigation. However, they found
the palpation test in the apical region to be significant in discriminating References
between pulpitis (symptomatic pulpal) and apical periodontitis (symp- 1. Seltzer S, Bender IB, Ziontz M. The dynamics of pulp inflammation: correlations
tomatic periapical). Nusstein and Beck (20) also reported similar find- between diagnostic data and actual histologic findings in the pulp. Oral Surg Oral
ings in a total of 323 patients. Med Oral Pathol 1963;16:846 –71.
Several symptoms previously believed to have differential diagnos- 2. Seltzer S, Bender IB, Ziontz M. The dynamics of pulp inflammation: correlations
tic power were found insignificant. Pain descriptors, such as intermit- between diagnostic data and actual histologic findings in the pulp. Oral Surg Oral
Med Oral Pathol 1963;16:969 –77.
tent, provoked, radiating, spontaneous, throbbing, localized, and dif- 3. Dummer P, Hicks R, Huws D. Clinical signs and symptoms in pulp disease. Int Endod
fused, did not show any significant correlations with the endodontic J 1980;13:27–35.
diagnoses. This was either because of an even distribution of these 4. Garfunkel A, Sela J, Ulmansky M. Dental pulp pathosis.
variables in different diagnostic groups or because of their rare occur- 5. Hayes R, Corcoran J, Zillich R. Diagnosis and treatment of diseases of the dental pulp
and periapical tissues. J Mich Dent Assoc 1981;63:313–7.
rence in the material. Klausen et al. (9) also noted insignificance rela- 6. Antonelli J. Acute dental pain, part II: diagnosis and emergency treatment. Compend
tive to diagnoses with a number of symptoms, such as spontaneous pain, Contin Educ Dent 1990;11:526 –33.
constant or intermittent pain, character of pain (sharp, dull, throbbing, 7. Carrotte P. Current practice in endodontics 2: diagnosis and treatment planning.
and irradiating), and duration of pain. These results underscore the Dent Update 2000;27:388 –91.
assumption that although pain descriptors remain useful and are per- 8. Steffel C. Clinical diagnosis in endodontics. J Indiana Dent Assoc 1986;65:22– 4.
9. Klausen B, Helbo M, Dabelsteen E. A differential diagnostic approach to the symp-
haps necessary diagnostic tools, pain descriptors themselves cannot tomatology of acute dental pain. Oral Surg Oral Med Oral Pathol 1985;59:297–301.
make the diagnosis alone. However, to keep things in proper perspec- 10. Morse DR, Seltzer S, Sinai I, Biron G. Endodontic classification. J Am Dent Assoc
tive, it must be remembered that this study tested the significance of 1977;94:685–9.
these pain descriptors in differential diagnosis of four uniquely grouped 11. Smulson MH. Classification and diagnosis of pulpal pathosis. Dent Clin North Am
1984;28:699 –723.
pulpal and periapical diagnostic categories. These pain descriptors still 12. Seltzer S. Classification of pulpal pathosis. Oral Surg Oral Med Oral Pathol
remain invaluable in differentiating between reversible and irreversible 1972;34:269 – 87.
pulpitis. This aspect of pain descriptors was not investigated in this 13. Shafer WG, Hine MK, Levy BM. A textbook of Oral Pathology. Ed 4. Philadelphia, PA:
article. Saunders, 1983.
It is possible that the odds ratios for variables may have been 14. Baume LJ. Diagnosis of diseases of the pulp. Oral Surg Oral Med Oral Pathol
1970;29:102–16.
overestimated because of bias. It could be a problem of circularity; for 15. De Quadros I, Gomes BP, Zaia AA, Ferraz CC, Souza-Filho FJ. Evaluation of endodontic
example, if the percussion test is the gold standard for diagnosing symp- treatments performed by students in a Brazilian dental school. J Dent Educ
tomatic periapical condition, and the clinician’s diagnosis is based, at 2005;69:1161–70.
least in part, on the results of this test, then this test will predominate the 16. Serene TP, Spolsky VW. Frequency of endodontic therapy in a dental school setting.
J Endod 1981;7:385–7.
diagnostic group. Such an apparently circular procedure represents a 17. Bender IB. Pulpal pain diagnosis—a review. J Endod 2000;26:175–9.
limitation but is unavoidable in this type of study because of the absence 18. Saad AY, Clem WH. An evaluation of etiologic factors in 382 patients treated in a
of a true “gold standard.” postgraduate endodontic program. Oral Surg Oral Med Oral Pathol 1988;65:91–3.
In conclusion, caries was more likely associated with symptomatic 19. Narhi M, Jyvasjarvi E, Virtanen A, Huopaniemi T, Ngassapa D, Hirvonen T. Role of
pulpal conditions and the main causative factor for root canal treat- intradental A- and C-type nerve fibres in dental pain mechanisms. Proc Finn Dent Soc
1992;88(suppl 1):507–16.
ments. A higher percentage of patients with painful pulpal conditions 20. Nusstein JM, Beck M. Comparison of preoperative pain and medication use in emer-
seek emergency care. Sharp pain was more likely associated with pulpal gency patients presenting with irreversible pulpitis or teeth with necrotic pulps. Oral
pathology, whereas dull pain was more likely associated with periapical Surg Oral Med Oral Pathol Oral Radiol Endod 2003;96:207–14.

JOE — Volume 33, Number 5, May 2007 Pulp and Periapical Pain 551