1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
PainDETECT: A suitable screening tool of neuropathic pain in
patients with painful post traumatic trigeminal nerve injuries?
Leigh-Ann Elias
Department of Oral Surgery, King’s College London Dental Institute, London, United Kingdom
Zehra Yilmaz
Department of Oral Surgery, King’s College London Dental Institute, London, United Kingdom
Jared G. Smith
Section of Mental Health, Division of Population Health Sciences & Education, St George's,
University of London, London, United Kingdom
Miriam Bouchiba
King’s College London, London, United Kingdom
Ruben A van der Valk
King’s College London, London, United Kingdom
Lisa Page
Department of Oral Surgery, King’s College London Dental Institute, London United Kingdom /
Sussex Partnership NHS Foundation Trust, England, United Kingdom
Sarah Barker
Department of Oral Surgery, King’s College London Dental Institute, London United Kingdom
Tara Renton
Department of Oral Surgery, King’s College London Dental Institute, London, United Kingdom
Correspondence to:
Professor Tara Renton
Professor of Oral Surgery
King’s College London Dental Institute
Denmark Hill Campus, Bessemer Road, London, SE5 9RS, United Kingdom
Email: tara.renton@kcl.ac.uk
Fax: 0203 299 2313
Abstract
1
46
The painDETECT questionnaire (PD-Q), originally developed and validated in a multicentre
47
study of neuropathic pain (NeP) patients with back pain, is increasingly being applied to other
48
pain conditions. The present study assessed whether the PD-Q would be a suitable screening tool
49
for detecting NeP in patients with post-traumatic inferior-alveolar nerve injury (IANI) and
50
lingual nerve injury (LNI).
51
neuropathy were given the PD-Q at their clinic appointment or sent to them after their
52
consultation. Eighty-nine patients (IANI = 56, LNI = 33) were included in the study, 75 of whom
53
suffered from painful neuropathy. Of these patients who completed the questionnaire so as to
54
allow a summary score to be calculated (n = 56), 34% were classified as having ‘likely NeP’
55
according to the PD-Q, with 41% of patients scoring in the uncertain classification range and the
56
remaining quarter in the ‘likely nocicpetive’ classification. There was a significant association
57
between PD-Q scores and pain intensity levels across the sample with those classified as likely
58
NeP reporting high levels of pain. The results suggest that PD-Q in its current format is not a
59
suitable screening tool for NeP associated with IANI or LNI.
60
Key words: painDETECT; trigeminal neuropathic pain; lingual nerve; inferior alveolar nerve;
61
post-traumatic sensory neuropathy
A prospective cohort of patients with clinically diagnosed
62
2
63
Introduction
64
Chronic pain (pain of more than three months’ duration) can be a lasting problem for patients
65
with iatrogenic inferior-alveolar nerve injury (IANI), or lingual nerve injury (LNI). The most
66
common causes of these injuries include third molar surgery, routine exodontia, complications of
67
root canal treatment, and the placement of dental implants1-2. A large proportion of these
68
patients’ injuries give rise to symptoms with a neuropathic component, as indicated by
69
complaints of burning pain, allodynia, tingling and/or paraesthesia that are uncontrollable by
70
non-steroidal anti-inflammatory drugs1. These symptoms are debilitating and have substantial
71
repercussions, such as the emotional impact of pain and long term sensory deficit, which may
72
directly affect individuals’ coping mechanisms and responses to stress, distress, anxiety,
73
depression, treatment expectations and motivation to improve2-3.
74
Still a matter of debate, the International Association for the Study of Pain (IASP) defines
75
neuropathic pain (NeP) as ‘pain initiated or caused by a primary lesion or dysfunction in the
76
nervous system4. While increasing evidence supports abnormal central pain processing in
77
chronic cases5, diagnosis of NeP remains clinical, based on a characteristic symptom profile
78
(e.g., pins and needles, electric shock like sensations), somatosensory abnormalities (e.g.
79
hyperalgesia, hypoesthesia, allodynia) and, sometimes, ancillary tests6-7. The PainDETECT
80
Questionnaire (PD-Q) screening tool was developed with a view to simplify the diagnosis of NeP
81
for clinicians namely because it does not require a clinical examination8-9. It was originally
82
developed and validated in German in a multicentre study of chronic low back pain patients and
83
appears to be a reliable screening tool in this diagnostic group: it has sensitivity of 85%,
84
specificity of 80%, and positive predictive value of 83%9. The PD-Q has subsequently been
85
translated into 22 languages with various forms of delivery (for a review, see Bennett et al.8), and
3
86
applied in studies of sensory profiles of various patient populations, including those with
87
established
88
musculoskeletal pain12 , patients suffering from osteoarthritis (OA)13-14, and in patients with post-
89
surgical pain15.
neuropathic
pain
conditions10-11,
patients
reporting
widespread
chronic
90
In orofacial pain conditions, the diagnosis of NeP remains challenging to diagnose,
91
predominantly because of the absence of associated clinical and radiographic abnormalities16.
92
Although confirmation of NeP in orofacial pain conditions can be made with some existing
93
neurological tests, these tests have reduced accuracy in identifying subtle neuronal abnormalities
94
and tend to be costly2,17. Further, quantitative sensory testing (QST), which depends on
95
expensive equipment and is time consuming, is not always helpful in the differential diagnosis;
96
and QST abnormalities cannot be taken as a conclusive demonstration of neuropathic pain18-19.
97
Thus, there is a significant need in this clinical population for efficient screening methods or
98
tools providing a systematic approach to assessing NeP. A simple questionnaire-based measure,
99
such as the PD-Q, has the potential to alleviate some of the financial burden and impact of
100
persistent pain by early identification in neuropathic conditions and thus facilitate more timely
101
expedition of appropriate therapy.
102
In line with the further development of effective screening tools that can be easily utilised in
103
the clinical arena in an accurate and cost-effective manner, the rationale for this study was to
104
explore for the first time, whether or not the PD-Q is a useful tool for identifying NeP elements
105
in two different known nerve injuries of the orofacial region.
106
Materials and methods
107
Design
4
108
This was an observational clinical study assessing the suitability of using the painDETECT
109
questionnaire (PD-Q)9 to screen for neuropathic pain components amongst patients with
110
iatrogenic nerve injury.
111
Participants
112
A prospective cohort of patients with clinically diagnosed neuropathy were given the PD-Q at
113
their clinic appointment or sent to them after their consultation. Patients were included in the
114
study if they presented to the clinic with reported sensory changes due to iatrogenic inferior
115
alveolar nerve injury (IANI) or lingual nerve injury (LNI), and could read and write in English
116
sufficiently well to complete the questionnaire. Clinical diagnoses of neuropathic pain was
117
ascertained by obtaining a pain history from each patient, as well as confirmation of allodynia,
118
hyperalgesia or spontaneous neuropathic pain by several clinical neurosensory tests, as well as
119
by recording the fact that their pain was unresponsive to non-steroidal anti-inflammatory drugs.
120
The pain was recorded within a proven area of focal neuropathy corresponding to the trigeminal
121
nerve branch damaged by the surgical intervention. All patients, prior to completing the PD-Q,
122
underwent a trigeminal nerve examination carried out by the principal investigator (which
123
included a series of neurosensory tests1), whereby trigeminal nerve injury was confirmed and
124
neuropathy was diagnosed. Patients were excluded if they had chronic orofacial pain caused by
125
other conditions. Ethical approval for the study was provided by the London – London Bridge
126
Local NHS Research Ethics Committee (REC number 08/H0808/105).
127
The painDETECT Questionnaire (PD-Q)
128
The PD-Q included 11-point numerical rating scales (NRS) dedicated to the evaluation of a
129
patient’s reported current pain level and its strongest and average levels during the past month.
130
The PD-Q also contained nine other items, of which seven related to sensory responses and two
5
131
to the temporal and spatial characteristics of the pain pattern. By rating the seven items from
132
never (0) to very strongly (5) on a category scale and summing these with the scores for temporal
133
(-1 to +1) and spatial characteristics (0 or +2), a summary score (minimum -1, maximum 38) was
134
obtained. A total score of < 13 indicated that a neuropathic component was unlikely, whereas a
135
score of > 18 indicated that a neuropathic component was likely. The online version of the
136
questionnaire is available at: http://www.virtualmedicalcentre.com/calc_pfizer_pain_detect.asp
137
Statistical analysis
138
Summary statistics for the overall sample and IANI and LNI patient subgroups were calculated
139
and presented in the form of means and standard deviations for quantitative variables and
140
frequency and percentages for qualitative variables. Differences between the two diagnostic
141
subgroups on pain-related and PD-Q indicators were measured using one-way analysis of
142
variance (ANOVA) or non-parametric equivalent for quantitative variables (according to data
143
distribution) and χ2 for qualitative variables. Associations between PD-Q NeP classifications and
144
measures of pain intensity (now, 4-week strongest, and 4-week average strength) were evaluated
145
considering the whole sample by one-way ANOVA models with post-hoc (pairwise)
146
examinations for significant results. To assess for correlations between the total PD-Q score and
147
4-week average pain intensity in each of the IANI and LNI groups, Spearman's rank-coefficient
148
tests were used. The criterion for statistical significance was set at P < 0.05, with no adjustments
149
for multiple comparisons given the descriptive nature of the study. All statistical analyses were
150
completed with the Statistical Package for the Social Sciences, Release 19.0 (SPSS, IBM).
151
6
152
Results
153
Sample characteristics
154
Altogether, 91 patients completed the PD-Q, 16 at their clinic appointment and 75 via
155
questionnaires posted to them after their consultation. Two questionnaires were completed
156
anonymously so these patients’ data could not be included in analyses. Eighty-nine iatrogenic
157
nerve injury patients were therefore included in the study; 56 with inferior alveolar nerve injury
158
(IANI) and 33 with lingual nerve injury (LNI). Most patients were female (68.5%). The mean
159
age of participants was 44.26 years ± (SD) 13.57 (range = 24 to 85), although IANI patients
160
tended to be older than their LNI counterparts (IANI = 47.02 ± 13.64; LNI = 39.58 ± 12.26; t(87)
161
= 2.58, P = 0.012). Mean duration of nerve injury was 18.50 months ± 28.50 (range = 1 to 216),
162
with the majority sustaining injury more than 6 months prior to the study commencing (n = 54 or
163
59%; IANI = 35 and LNI = 19). A little more than half of all patients’ nerve injury was sustained
164
during third molar surgery (TMS: 51.7%), although this was more likely for the LNI (81.8%)
165
than IANI (33.9%, P < 0.001) patients, with IANI arising from a wider range of procedures
166
(Table 1).
167
A minority of patients were receiving one or more prescription medications at the time of the
168
study (19 or 21.3%; IANI = 14 or 25.0%, LNI = 5 or 15.2%), six (5 IANI, 1 LNI) who were
169
taking two or more different medication classes. Of those receiving medications, 9 (IANI = 7,
170
LNI = 2) were taking antiepileptic medications (pregabalin, gabapentin, carbamazepine or
171
oxcarbazepine), 9 (IANI = 7, LNI = 2) were receiving analgesics (lidocaine, benzocaine,
172
tramadol or paracetamol), 8 (IANI = 5, LNI = 3) were taking antidepressants (tricyclic
173
antidepressants or selective serotonin reuptake inhibitors), one (LNI) taking a tranquilizer
174
(benzodiazepine), and another (IANI) on a course of cortisone. Seven patients (7.9%; IANI = 1
7
175
or 1.8%, LNI = 6 or 18.2%) had previously undergone surgery intended to resolve their nerve
176
injury.
177
INSERT TABLE 1 ABOUT HERE
178
179
180
Pain intensity
181
Patients indicated low levels of pain using the visual analogue scale (VAS) scores. Notably, 12
182
patients (7 IANI, 5 LNI) reported a zero average pain in the last 4 weeks, indicating that they had
183
not recently experienced pain as part of their condition, while 2 patients did not respond to the
184
average pain intensity item. For those patients reporting pain, mean 4-week average strength of
185
pain was moderate for both IANI (4.17 ± 2.56) and LNI patients (4.46 ± 2.70). The mean pain
186
intensity at time of questionnaire completion was 3.66 ± 2.71 for IANI patients and 3.93 ± 3.11
187
for LNI patients, while mean strongest pain levels reported within the previous 4 weeks were
188
5.49 ± 2.87 for IANI patients and 6.15 ± 2.92 for LNI patients. Reported pain levels were not
189
significantly different between the IANI and LNI patient groups (for all comparisons, P >
190
0.340). Pain levels (4-week average) were not related to whether patients were receiving
191
medication for pain (anti-epileptic and/or analgesic medication; n = 14; mean = 4.43 ± 2.50) or
192
not (n = 61; mean = 4.25 ± 2.64; P = 0.752) or if patients had undergone surgery for nerve injury
193
(n = 7; mean = 4.86 ± 2.61) or had not (n = 68; mean = 4.22 ± 2.61; P = 0.532).
194
When considering all patients who reported pain (secondary to their nerve injury), the most
195
frequently selected pain patterns were ‘persistent pain with slight fluctuations’ (46.5%) and pain
196
attacks characterized by pain-free intervals (28.2%; Fig 1). Although LNI patients tended to
197
report their course of pain as being persistent with pain attacks more than patients with IANI
8
198
(25.9% vs. 11.4%), and conversely, IANI patients more often described their pain pattern as pain
199
attacks with pain between them than did LNI patients, there was no significant difference in the
200
distribution of pain type choices between IANI and LNI patient groups (2(3) = 3.38, P = 0.337).
201
Of note, 13 patients experiencing pain did not respond to the item concerning the presence of
202
radiating pain and 4 failed to indicate the pattern of their pain.
203
204
INSERT FIGURE 1 ABOUT HERE
205
206
Frequency of sensory symptoms
207
Seven questions of the PD-Q addressed the quality and intensity of specific neuropathic
208
symptoms, namely burning, prickling, allodynia, attacks, thermal sensitivity, numbness and
209
pressure. The patients could rate the perceived severity of each of these symptoms from 0-5
210
(never, hardly noticed, slightly, moderately, strongly, very strongly). Figure 2 shows the
211
frequency of the sensory disturbances that were regarded as clinically relevant (i.e., if the
212
patients marked a score of > 3 [strongly or very strongly]).
213
214
INSERT FIGURE 2 ABOUT HERE
215
216
IANI and LNI patients indicated clinically relevant numbness at comparably high
217
frequencies. Sixteen (28.6%) patients with IANI indicated clinically relevant allodynia; a
218
proportion that was significantly greater than that reported by LNI patients. Likewise, clinically
219
relevant thermal sensitivity was indicated by 14 (25.5%) patients with IANI, as opposed to only
9
220
2 LNI patients. These results suggest that these sensory symptoms in the PD-Q were more
221
prevalent in IANI patients.
222
Detection of NeP using PD-Q in IANI and LNI patients experiencing pain
223
The purpose of the PD-Q is to classify whether a person’s (lower-back) pain is likely to have a
224
neuropathic or non-neuropathic component. Specifically, the PD-Q screens patients into three
225
pain groups: ‘likely nociceptive’ ‘unclear’, and ‘likely neuropathic’. The utility of the tool in its
226
original form was examined in those trigeminal nerve injury patients who reported pain as part of
227
their condition (n = 75). Unfortunately, a number of patients (19 or 25.3%; IANI = 12 or 25.5%;
228
LNI = 7 or 25.0%) did not complete all items necessary to calculate a final score on the PD-Q
229
(predominantly because they failed to response to either one or both items concerning the
230
presence of radiating pain and the pain pattern). Non-completers were not distinguishable from
231
those who completed the PD-Q in any way. Questionnaire completion was not linked to reported
232
pain levels, with 4-week average intensity scores comparable for non-completers (n = 19; mean
233
= 4.42 ± 2.52) and completers (n = 56; mean = 4.23 ± 2.64; P = 0.731), and no demographic or
234
nerve-injury related variable predicted non-completion of the PD-Q.
235
236
INSERT TABLE 2 ABOUT HERE
237
238
The mean PD-Q scores and associated NeP classifications of the 56 PD-Q completers are
239
shown in Table 2. Patients with IANI evidenced higher PD-Q scores than LNI patients, although
240
the group difference only approached significance. Importantly, only 19 (33.9%) of all patients
241
were classified in the ‘likely neuropathic’ category. Twenty-three (41.1%) patients scored in the
242
midrange in which NeP classification is reportedly uncertain, while the remaining quarter (14)
10
243
fell in the ‘unlikely NeP’ classification. Compared with LNI patients, a greater proportion of
244
IANI patients scored above the NeP cut-off and a lower proportion evidenced scores indicating
245
unlikely NeP, although none of the proportion differences were significant. In summary, the
246
questionnaire failed to identify all patients with NeP associated with post traumatic trigeminal
247
neuropathy in this patient cohort.
248
Relationship between PD-Q scores and pain severity
249
A number of previous studies have reported positive associations between pain intensity and PD-
250
Q scores.9,10,14 The relationship between pain severity and the PD-Q results was first investigated
251
by dividing all nerve injury patients into three groups in accordance with Freynhagen et al.’s9
252
classification system and comparing intensity scores on all three pain measures. For all pain
253
measures, there was a clear linear trend on increasing severity from those scoring less than 13 on
254
the PD-Q, those scoring 13-18, and those with total scores that were greater than 18 (Fig 3).
255
Kruskal-Wallis tests demonstrated a significant difference among the three classification groups
256
of patients for strongest pain during the last four weeks (P = 0.008) and average pain intensity
257
during the past four weeks (P = 0.032), but differences in pain intensity now narrowly missed
258
significance (P = 0.069). Post hoc (pairwise) analyses showed that patients scoring more than 18
259
on the PD-Q had significantly higher pain severity scores than did patients scoring less than 13
260
across all measures (for all comparisons, P < 0.042), while differences between those scoring
261
greater than 18 and those scoring between 13 and 18 were significant for strongest pain only (P
262
= 0.011). There were no significant differences between patients scoring between 13 and 18 and
263
those scoring less than 13 (for all comparisons, P > 0.050). Correlational analyses using
264
continuous questionnaire scores also revealed a significant relationship between PD-Q scores
265
and reported 4-week average pain intensity (rs(55) = 0.33, P = 0.012). Figure 4 shows a positive
11
266
correlation between PD-Q scores and average pain intensity in both IANI and LNI patient
267
groups, although this relationship was significant only in the former group, (rs(34) = 0.39, P =
268
0.022), and not the latter which was weaker (rs(34) = 0.24, P = 0.306).
269
INSERT FIGURE 3 ABOUT HERE
270
271
272
Notably, PD-Q sensitivity in TNI patients improved by applying a relevant threshold based
273
on reported pain severity10, although the considered sample diminished. For example, if PD-Q
274
classifications were made for only those patients reporting (4-week average) pain intensity of ≥ 4
275
(n = 29) then those classified with ‘likely neuropathic’ pain increased to 44.8%. Raising the
276
threshold further, to a reported pain intensity of ≥ 6 (n = 20), yielded 55% of patients in the
277
‘likely neuropathic’ classification, and another 35% in the ‘uncertain’ classification - thus only
278
10% falling in the ‘unlikely NeP’ classification. Of note, there were no differences across PD-Q
279
classification groups with respect to either age (P = 0.327) or duration of injury (P = 0.351) and
280
no observed relationship between classification groups and gender (P = 0.845), cause of injury
281
(TMS vs other; P = 0.156), or whether or not patients were receiving medication for pain (P =
282
0.338).
283
284
INSERT FIGURE 4 ABOUT HERE
285
286
Discussion
287
For nerve injury patients experiencing pain, the painDETECT questionnaire’s sensitivity in
288
detecting neuropathic trigeminal pain in this study was 34%. Notably, 24% of LNI patients with
12
289
pain were classified as having ‘likely neuropathic’ pain from the PD-Q compared to 40% of
290
IANI patients with pain, a (numerical) difference attributable to the latter’s greater propensity to
291
suffer from allodynia and sensitivity to thermal stimuli1. To the extent that patients in this study
292
were reporting pain secondary to their nerve injury, and had been diagnosed with neuropathy as a
293
result of the injury, it is highly likely their pain has a significant neuropathic component. The low
294
rate of classification of neuropathic pain using the PD-Q (originally developed and validated in
295
patients with back pain) therefore suggests that the use of the PD-Q as a tool to detect
296
neuropathic pain in the TNI patient sample is not appropriate in its current form.
297
Despite the apparent low sensitivity of the PD-Q, there was a modest but significant
298
correlation between PD-Q scores and 4-week average strength of pain, although this relationship
299
was more obvious in the group of patients with IANI. Across all patients, comparisons of pain
300
intensity levels between NeP classification groups indicated significantly higher NRS pain scores
301
amongst patients with PD-Q scores of greater than 18, thus showing some advantage of using the
302
PD-Q to screen for NeP components in TNI patients who suffer from higher levels of pain.
303
Previous studies examining the utility of the PD-Q in various chronic pain populations have
304
adopted thresholds at the outset for pain intensity10,
20
305
average pain intensity levels for TNI patients’ (mean = 4.3, SD = 2.6) were relatively low in
306
comparison with those found in other NeP populations such as patients with painful diabetic
307
peripheral neuropathy11. When thresholds for pain intensity of 4 and 6 (on a 0-10 NRS) were
308
applied to the sample here, those classified with ‘likely neuropathic’ pain increased (from
309
33.9%) to 44.8% and 55.0%, respectively, a notable improvement although still indicative of
310
poor sensitivity. However, as almost half the patients with painful neuropathy secondary to TNI
. In the present study, reported 4-week
13
311
in this study report pain levels that average less than 4, applying these thresholds is likely to be
312
of limited value.
313
Only one other study has assessed the utility of the PD-Q in orofacial pain. Ukwas et al.21
314
evaluated the diagnostic ability and reproducibility of the PD-Q in a sample of patients suffering
315
from orofacial pain (OFP). They reported that the PD-Q had limited accuracy in detecting NeP
316
(as measured against clinical diagnosis) and yielded low reproducibility over time. These
317
findings are broadly consistent with the present study, in which the PD-Q showed poor
318
sensitivity in the identification of neuropathic elements in OFP associated with post-traumatic
319
neuropathy, and suggest that the validity and reliability PD-Q in OFP is not satisfactory.
320
Although the PD-Q may suggest the presence of NeP components in OFP, these are not typical
321
NeP lesions, and may be considered a distinct subtype with unique patterns of
322
symptomatology22. As such, we cannot assume that those patients with post traumatic trigeminal
323
neuropathy scoring 18 or less on the PD-Q do not experience neuropathic pain.
324
Suggested improvements to PD-Q in orofacial pain
325
Amendment of the PD-Q is therefore necessary, not only to better detect (trigeminal) NeP, but
326
also to improve compliance with answering all questions. The modest completion rate of the PD-
327
Q (only 56 of the 75 patients experiencing pain responded to all PD-Q items) complicates
328
interpretation of findings in this cohort and suggests that the face validity of the standard PD-Q is
329
low in TNI patients. In this instance, missing or invalid responses likely reflect the lack of
330
relevance of specific items for TNI patients. Suggested changes include image modification; the
331
PD-Q image could depict simply the head and neck region including inside the oral cavity
332
(tongue, gingivae and floor of the mouth). The patients also may not have a full understanding of
333
“radiating” pain corresponding to the original pain map from the Freynhagen et al.9 study. The
14
334
image would represent the orofacial region on a larger scale and may help to avoid confusion
335
with pain from other areas. Modification to items pertaining to activity-elicited pain is likely to
336
be beneficial also. For example, if the question concerning thermal stimuli; “Is cold or heat (bath
337
water) in this area occasionally painful?”, is modified to; “When washing your face with cold or
338
hot water, is this area occasionally painful?”, more patients with orofacial pain may indicate the
339
presence of neuropathy. Likewise, inclusion of a question related to pain upon having a hot/cold
340
drink or food is likely to be more relevant to patients with orofacial NeP (particularly those with
341
LNI). Recently, Hochman and colleagues13 modified items in the PD-Q, framing questions to ask
342
about symptoms ‘in or around’ the worst affected knee, to better assess the NeP elements in a
343
cohort of patients with knee osteoarthritis (OA). The modified questionnaire was successful in
344
identifying a quarter of OA patients with NeP symptoms localised to the knee, and suggested to
345
be a clinically valid tool to determine NeP in adult knee OA20. A similar approach, with a focus
346
on relating identified neuropathic symptoms to those neurosensory changes found at clinical
347
assessment, may yield dividends in developing a valid questionnaire-based tool to detect NeP in
348
orofacial pain.
349
Limitations
350
A small number of patients in the study were undergoing pharmacological treatment for their
351
condition at the time of PD-Q administration. While there were no differences in PD-Q scores
352
between those receiving medication and those who did not, treatment effects cannot be entirely
353
ruled out. Also, questionnaires were administered to TNI patients at consultation or later (via
354
post), and there is the possibility that responses could have been influenced by the clinical
355
consultation procedure, and/or, by other members of the household. More generally, patients
356
with iatrogenic trigeminal nerve injuries often suffer from multiple symptoms that include NeP,
15
357
numbness and altered sensation1,23, and this mixture of (related) symptoms may have confused
358
patients’ responses (with some patients interpreting the terms interchangeably). Patients may
359
have benefitted from completing the questionnaire within the clinic so that the clinician can
360
provide aid when questions are not understood. However, the PD-Q was intended to be a self-
361
administered measure and purported to be easily understood and filled in by the patient 9-10 . The
362
modest compliance rate in the present study suggests modification to specific items is warranted
363
rather than changes in PD-Q administration. Finally, there are a range of psychological factors
364
known to promote NeP symptoms like catastrophising24, anxiety and depression25-26, and the
365
influence of these phenomena on PD-Q responding was not assessed here.
366
Conclusions
367
Developing a simple and inexpensive method for screening of NeP in patients with orofacial pain
368
conditions is likely to aid decisions about patient management and treatment strategies in a range
369
of dental and oral health care settings. This study showed that the PD-Q tool evidenced limited
370
sensitivity to the neuropathic component to pain experienced by a sample of patients’ diagnosed
371
with neuropathy secondary to trigeminal nerve injury. This suggests the questionnaire, in its
372
present form, is not an adequate screening measure in patients with orofacial pain. A PD-Q that
373
is specifically modified for use in orofacial pain may better facilitate the identification of NeP in
374
this group of patients.
375
376
16
377
References
378
1.
379
380
Renton T, Yilmaz Z. Profiling of patients presenting with posttraumatic neuropathy of the
Trigeminal Nerve. J Orofac Pain 2011: 25: 333-4.
2.
Rodríguez-Lozano F, Sanchez-Pérez A. Neuropathic orofacial pain after dental implant
381
placement: review of the literature and case report. Oral Surg Oral Med Oral Pathol Oral
382
Radiol Endod 2010: 109: e8-12.
383
3.
384
385
Jerjes W, Hopper C, Kumar M, Upile T, Madland G, Newman S, Feinmann C.
Psychological intervention in acute dental pain: review. Br Dent J 2007: 202(6): 337-43.
4.
Treede R, Jensen T, Campbell JN, Cruccu G, Dostrovsky JO, Griffin JW, Hansson P,
386
Hughes R, Nurmikko T, Serra J. Neuropathic pain: Redefinition and a grading system for
387
clinical and research purposes. Neurology 2008: 70: 1630-5.
388
5.
389
390
Kehlet H, Jensen TS, Woolf CJ. Persistent postsurgical pain: Risk factors and prevention.
Lancet 2006: 367: 1618-25.
6.
Cruccu G, Sommer C, Anand P, Attal N, Baron R, Garcia-Larrea L, Haanpaa M, Jensen
391
TS, Serra J, Treede RD. EFNS guidelines on neuropathic pain assessment: revised 2009.
392
Eur J Neurol 2010: 17(8): 1010-8.
393
7.
394
395
Gilron IC, Watson PN, Cahill CM, Moulin DE. Neuropathic pain: a practical guide for the
clinician. CMAJ 2006: 175(3): 265-75.
8.
Bennett MI, Attal N, Backonja MM, Baron R, Bouhassira D, Freynhagen R, Scholz J,
396
Tölle TR, Wittchen HU, Jensen TS. Using screening tools to identify neuropathic pain.
397
Pain 2007: 127(3): 199-203.
17
398
9.
Freynhagen R, Baron R, Gockel U, Tölle TR. pain DETECT: a new screening
399
questionnaire to identify neuropathic components in patients with back pain. Curr Med Res
400
Opin 2006: 22(10): 1911-20.
401
10.
De Andres J, Perez-Cajaraville J, Lopez-Alarcon MD, Lopez-Millan JM, Margarit C,
402
Rodrigo-Royo MD, Franco-Gay ML, Abejon D, Ruiz MA, Lopez-Gomez V, Perez M.
403
Cultural adaptation and validation of the painDETECT scale into Spanish. Clin J Pain
404
2012: 28: 243-53.
405
11.
Koroschetz J, Rehm SE, Gockel U, Brosz M, Freynhagen R, Tölle TR, Baron R.
406
Fibromyalgia and neuropathic pain--differences and similarities. A comparison of 3057
407
patients with diabetic painful neuropathy and fibromyalgia. BMC Neurol 2011: 11: 55.
408
12.
Amris K, Jespersen A, Bliddal H. Self-reported somatosensory symptoms of neuropathic
409
pain in fibromyalgia and chronic widespread pain correlate with tender point count and
410
pressure-pain thresholds. Pain 2011: 151: 664–9.
411
13.
412
413
Hochman JR, Gagliese L, Davis AM, Hawker GA. Neuropathic pain symptoms in a
community knee OA cohort. Osteoarthritis Cartilage. 2011: 19(6): 647-54.
14.
Shigemura T, Ohtori S, Kishida S, Nakamura J, Takeshita M, Takazawa M, Harada Y,
414
Takahashi K. Neuropathic pain in patients with osteoarthritis of hip joint. Eur Orthop
415
Traumatol 2011: 2: 73-7.
416
15.
417
418
419
Wylde V, Hewlett S, Learmonth ID, Dieppe P. Persistent pain after joint replacement:
prevalence, sensory qualities and postoperative determinants. Pain 2011: 152: 566–72.
16.
Vickers ER, Cousins MJ. Neuropathic Orofacial Pain Part 2–Diagnostic Procedures,
Treatment Guidelines and Case Reports. Aust Endod J 2000: 26(2): 53-63.
18
420
17.
Svensson P, Baad-Hansen L, Pigg M, List T, Eliav E, Ettlin D, Michelotti A, Tsukiyama
421
Y, Matsuka Y, Jaaskelainen SK, Essick G, Greenspan JD, Drangsholt M. Guidelines and
422
recommendations for assessment of somatosensory function in orofacial pain conditions –
423
a taskforce report. J Oral Rehabil 2011: 38(5): 366-94.
424
18.
425
426
Cruccu G, Anand P, Attal N, Garcia-Larrea L, Haanpää M, Jørum E, Serra J, Jensen TS.
EFNS guidelines on neuropathic pain assessment. Eur J Neurol 2004: 11(3): 153-62.
19.
Pfau DB, Geber C, Birklein F, Treede RD. Quantitative sensory testing of neuropathic pain
427
patients: potential mechanistic and therapeutic implications. Curr Pain Headache Rep
428
2012: 16(3): 199-206.
429
20.
Morso L, Kent PM, Albert HB. Are self-reported pain characteristics, classified using the
430
PainDETECT questionnaire, predictive of outcome in people with low back pain and
431
associated leg pain? Clin J Pain 2011: 27: 535-41.
432
21.
433
434
Ukwas A, Tonks S, Zakrzewska J, Leeson R. An exploration of the Pain Detect
Questionnaire as a screening tool in Oro-facial pain. British J Pain 2012: 6: 24.
22.
Attal N, Fermanian C, Fermanian J, Lanteri-Minet M, Alchaar H, Bouhassira D.
435
Neuropathic pain: are there distinct subtypes depending on the aetiology or anatomical
436
lesion? Pain 2008: 138: 343–53.
437
23.
438
439
Renton T, Adey-Viscuso D, Meechan JG, Yilmaz Z. Trigeminal nerve injuries in relation
to the local anaesthesia in mandibular injections. Br Dent J 2010: 209(9): E15.
24.
Sullivan MJ, Lynch ME, Clark AJ. Dimensions of catastrophic thinking associated with
440
pain experience and disability in patients with neuropathic pain conditions. Pain 2005:
441
113(3): 310-5.
19
442
25.
443
444
Haythornthwaite JA, Clark MR, Pappagallo M, Raja SM. Pain coping strategies play a role
in the persistence of pain in post-herpetic neuralgia. Pain 2003: 106: 453-60.
26.
Turk DC, Audette J, Levy RM, Mackey SC, Stanos S. Assessment and treatment of
445
psychosocial comorbidities in patients with neuropathic pain. Mayo Clin Proc 2010: 85:
446
S42-50.
447
448
Figure Legends
449
Fig 1 Pain pattern by nerve injury: IANI = inferior alveolar nerve injury; LNI = lingual nerve
450
injury.
451
Fig 2 Frequency (percentage of patients) indicating clinically relevant problems (i.e., score > 3)
452
on sensory items in the PainDETECT questionnaire (PD-Q). Note: IANI = inferior alveolar
453
nerve injury; LNI = lingual nerve injury. * Indicates significant differences between groups (P <
454
0.05). Please note: the n values for each sensory item are slightly variable due to a small number
455
of patients failing to respond on each item (between 2 and 4 responses missing for each item).
456
Fig 3 Association between intensity measures of pain, and neuropathic pain classification
457
according to total score on the PainDETECT Questionnaire (PD-Q). Strongest pain and average
458
pain were reported for the past four weeks. Mean scores are shown for each PD-Q classification
459
(PD-Q<13, n = 14; PD-Q 13-18, n = 23; PD-Q>18, n = 19). Error bars represent the standard
460
error of the mean. Pairwise group comparisons were administered for each measure: an asterisk
461
(*) indicates a significant difference from the PD-Q< 13 group, a double asterisk (**) represent
462
significant differences from PD-Q<13 and PD-Q 13-18 groups.
463
Fig 4 Relationship between PainDETECT questionnaire (PD-Q) scores and reported average
464
strength of pain during the past four weeks for patients with inferior alveolar nerve injury (n =
465
35) and lingual nerve injury (n = 21).
20
466
Table 1: Aetiology of Injury in Patients with Inferior Alveolar Nerve Injury (IANI) and Lingual
467
Nerve Injury (LNI)
468
Questionnaire
Third molar surgery (TMS)
Implant placement under LA
Chemical injury from the LA
Extraction of mandibular tooth/teeth
(apart from TMS )
Restorative (e.g., endodontic) treatment
Trauma (non-surgical; e.g., assault)
Pathological excision
Apicectomy
Other
Unknown
IANI (n = 56)
n
%
LNI (n = 33)
n
%
19
12
5
33.9
21.4
8.9
27
81.8
2
6.1
9
16.1
2
6.1
3
2
1
1
4
5.4
3.6
1.8
1.8
7.1
1
1
3.0
3.0
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
21
484
Table 2. Detection of NeP Using PainDETECT Questionnaire (PD-Q) in Inferior Alveolar
485
Nerve Injury (IANI) and Lingual Nerve Injury (LNI) Patients Experiencing Pain
486
PainDETECT Questionnaire Scores
PD-Q score (-1 to 38; excludes those
reporting no pain in the last 4 weeks)
NeP Prevalence using cut-off scores
Number of people below neuropathic
pain cut-off (score  12)
Number of people indicating uncertain
neuropathic classification (score 13-18)
Number of people above neuropathic
cutoff (score ≥ 19)
IANI (n = 35)
LNI (n = 21)
IANI vs LNI
Mean±SD
Mean±SD
F
P
17.91±7.20
14.33±7.24
3.24
0.078
n (%)
n (%)
χ²
P
6 (17.1)
8 (38.1)
3.07
0.080
15 (42.9)
8 (38.1)
0.12
0.726
14 (40.0)
5 (23.8)
1.54
0.215
Notes: SD = standard deviation. Group comparison analysis for PD-Q scores and prevalence
using cut-off scores performed using one-way ANOVA and Chi-squared tests, respectively.
487
488
489
490
491
492
493
494
495
496
497
498
499
500
22
501
Figure 1: Pain pattern by nerve injury: IANI = inferior alveolar nerve injury; LNI = lingual nerve
502
injury.
503
504
505
506
507
508
509
510
511
512
513
23
514
Fig 2 Frequency (percentage of patients) indicating clinically relevant problems (i.e., score > 3)
515
on sensory items in the PainDETECT questionnaire (PD-Q). Note: IANI = inferior alveolar
516
nerve injury; LNI = lingual nerve injury. * Indicates significant differences between groups (P <
517
0.05). Please note: the n values for each sensory item are slightly variable due to a small number
518
of patients failing to respond on each item (between 2 and 4 responses missing for each item).
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
24
539
Fig 3 Association between intensity measures of pain, and neuropathic pain classification
540
according to total score on the PainDETECT Questionnaire (PD-Q). Strongest pain and average
541
pain were reported for the past four weeks. Mean scores are shown for each PD-Q classification
542
(PD-Q<13, n = 14; PD-Q 13-18, n = 23; PD-Q>18, n = 19). Error bars represent the standard
543
error of the mean. Pairwise group comparisons were administered for each measure: an asterisk
544
(*) indicates a significant difference from the PD-Q< 13 group, a double asterisk (**) represent
545
significant differences from PD-Q<13 and PD-Q 13-18 groups.
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
25
563
Fig 4 Relationship between PainDETECT questionnaire (PD-Q) scores and reported average
564
strength of pain during the past four weeks for patients with inferior alveolar nerve injury (n =
565
35) and lingual nerve injury (n = 21).
566
567
568
569
570
571
26