Accepted Manuscript

The efficacy of codeine when added to paracetamol (acetaminophen) and ibuprofen

for relief of postoperative pain after surgical removal of impacted third molars: a
double-blind randomised control trial

Adrian D. Best, BDSc, MB, ChB, BTh, MDS, FRACDS, R.K. De Silva, BDS,
FDSRCPS, FFDRCSI, FDSRCS, W.M. Thomson, BSc, BDS, MComDent, MA, PhD,
Darryl C. Tong, BDS, MB, ChB, MSD, PhD, FFDRCSI, FDSRCS, FACOMS, FACS,
MRSNZ, Claire M. Cameron, BSc, DipGrad, MSc, PhD, Harsha L. De Silva, BDS,
MS, FDSRCS, FFDRCS
PII: S0278-2391(17)30504-9
DOI: 10.1016/j.joms.2017.04.045
Reference: YJOMS 57798

To appear in: Journal of Oral and Maxillofacial Surgery

Received Date: 1 December 2016

Revised Date: 19 March 2017
Accepted Date: 27 April 2017

Please cite this article as: Best AD, De Silva RK, Thomson WM, Tong DC, Cameron CM, De Silva
HL, The efficacy of codeine when added to paracetamol (acetaminophen) and ibuprofen for relief of
postoperative pain after surgical removal of impacted third molars: a double-blind randomised control
trial, Journal of Oral and Maxillofacial Surgery (2017), doi: 10.1016/j.joms.2017.04.045.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED
The efficacy of codeine when added MANUSCRIPT
to paracetamol (acetaminophen) and ibuprofen for relief

of postoperative pain after surgical removal of impacted third molars: a double-blind

randomised control trial.

Adrian D. Best, BDSc, MB, ChB, BTh, MDS, FRACDS,1 R. K. De Silva, BDS, FDSRCPS, FFDRCSI,

PT
FDSRCS,2 W. M. Thomson, BSc, BDS, MComDent, MA, PhD,3 Darryl C. Tong, BDS, MB, ChB, MSD,

PhD, FFDRCSI, FDSRCS, FACOMS, FACS, MRSNZ,4 Claire M. Cameron, BSc, DipGrad, MSc, PhD,5

RI
Harsha L. De Silva, BDS, MS, FDSRCS, FFDRCS.6

SC
1. Former Oral and Maxillofacial Surgery Registrar, University of Otago, New Zealand.

U
2. Associate Professor, Oral and Maxillofacial Surgeon, University of Otago, New Zealand.
AN
3. Professor, Dental Epidemiologist, University of Otago, New Zealand.

4. Professor, Oral and Maxillofacial Surgeon, University of Otago, New Zealand.

M

5. Biostatistician, University of Otago, New Zealand.

D

6. Oral and Maxillofacial Surgeon, University of Otago, New Zealand.

TE
EP

Corresponding author details:

C

Dr. Adrian Best

AC

Mailing Address: c/o Assoc. Prof. Rohana K. De Silva

Department of Oral Diagnostic and Surgical Sciences.

Faculty of Dentistry, University of Otago.

310 Great King Street, Dunedin, 9016, New Zealand.

Phone: +64 3 479 7031

Email: adrianucx@gmail.com
 ACCEPTED
KEYWORDS: Dental. Pain. Wisdom MANUSCRIPT
teeth. Third molars. Postoperative. Analgesia. Non-

opioid. Paracetamol. Acetaminophen. Ibuprofen. Non-steroidal anti-inflammatory

drugs. Opioid. Codeine.

PT
Abstract

RI
Background: The use of opioids in combination with nonopioids is common practice for

acute pain management after third molar surgery. One such combination is

SC
paracetamol/ibuprofen/codeine. We assessed the efficacy of codeine when added to a

U
paracetamol/ibuprofen regimen for pain relief following third molar surgery.
AN
Methods: This study was a randomised, double-blind, placebo-controlled trial conducted
M

in patients undergoing the surgical removal of at least one impacted mandibular third
D

molar requiring bone removal. Participants were randomly allocated to either a control
TE

group (paracetamol 1000mg/ibuprofen 400mg) or an intervention group (paracetamol

1000mg/ibuprofen 400mg/codeine 60mg). All participants were treated under

EP

intravenous sedation and using identical surgical conditions and technique.

Postoperative pain was assessed using visual analogue scales every 3 hours (while
C

awake) for the first 48 hours following surgery. Pain was globally assessed using a
AC

questionnaire on day 3 after surgery.

Results: There were 131 participants (36% male), with 67 in the control group and 64 in

the intervention group. Baseline characteristics were similar for both groups. Data were

analysed using a modified intention-to-treat analysis and, for this, a linear mixed model

was used. The model showed that the baseline VAS score was associated with the
 ACCEPTED
subsequent VAS scores and that, MANUSCRIPT
with each 3hr period, the VAS score increased by an

average of 0.08. The treatment effect was not statistically significant, indicating there

was no difference in recorded pain levels between the two groups over the first 48 hours

following mandibular third molar surgery. Similarly, the two groups did not differ in

their global ratings of postoperative pain.

PT
Conclusion: Codeine 60mg added to a paracetamol 1000mg/ibuprofen 400mg regimen

RI
does not improve analgesia following third molar surgery.

U SC
AN
Introduction

Despite major advances being made in pain management, acute postoperative pain remains an
M

important clinical problem following third molar surgery. Patients undergoing outpatient

ambulatory third molar surgery commonly experience postoperative pain of moderate to

D

severe intensity, even when given optimal doses of oral opioid and or nonopioid analgesics.1-7
TE
EP

Using opioids in combination with nonopioids is a common strategy for managing acute pain

following third molar surgery. The rationale for this is that combinations of opioid and
C

nonopioid analgesics—which possess different mechanisms of action—produce greater

AC

analgesia than that produced by opioids or nonopioids alone.6 A frequently prescribed

opioid/nonopioid combination used postoperatively for third molar surgery is

ibuprofen/paracetamol/codeine. Ibuprofen (a non-steroidal anti-inflammatory drug) and

paracetamol (acetaminophen) are relatively safe and effective analgesics with minimal adverse

effects when prescribed and used correctly. By contrast, codeine (a weak opioid) has the

potential for significant adverse effects such as nausea, vomiting, constipation, sedation,
dizziness, seizures, respiratory ACCEPTED MANUSCRIPT
depression and even addiction. Moreover, some people are

unable to convert codeine into morphine, the agent responsible for codeine’s analgesic effect.8 9

Despite the negative characteristics of codeine, it remains a commonly prescribed analgesic

following third molar surgery. However, is it a necessary analgesic in this context? It may be

PT
beneficial to patients if it could be demonstrated that codeine is not needed. We investigated

the hypothesis that the addition of codeine 60mg 6 hourly to a regimen of paracetamol

RI
1000mg 6 hourly and ibuprofen 400mg 8 hourly does not improve analgesia following

SC
mandibular third molar surgery involving bone removal.

U
Methods
AN
This study was approved by the Health and Disability Ethics Committee, New Zealand (ethics

committee reference number: 13/STH/35). It was also approved by the University of Otago
M

Human Ethics Committee, New Zealand (ethics committee reference number: 13/080).
D
TE

The study was a double-blind randomised control trial conducted at the University of Otago,

New Zealand. Recruitment, surgery and follow-up took place between November 2013 and
EP

September 2014.
C

We included adults requiring the removal of at least one mandibular third molar for which
AC

bone removal was expected, with or without removal of one or more upper third molars.

Anticipated bone removal was considered necessary to ensure adequate postoperative pain

intensity among participants. All participants had to be medically fit for third molar surgery

under intravenous sedation.

 ACCEPTED
We excluded patients if they: were MANUSCRIPT
under 17 years old; had an existing medical condition

requiring a medication with analgesic properties which was not able to be ceased prior to and

for the duration of the study; needed to drive a motor vehicle or operate machinery within 48

hours following surgery; had a history of NSAID-sensitive asthma, peptic ulceration, bleeding

disorder, renal or hepatic impairment, cardiovascular disease, systemic lupus erythematosus,

PT
lactose intolerance, respiratory depression, chronic obstructive pulmonary disease, opioid

addiction, alcoholism, G6PD deficiency, hypersensitivity to morphine or benzodiazepines,

RI
phenylketonuria, myasthenia gravis, glaucoma, osteoporosis, or psychosis, or were currently

SC
pregnant or breastfeeding; were taking an anticoagulant, hepatic enzyme inducer, or central

nervous system depressant; were suffering from systemic viral, bacterial or fungal infection; or

U
if the clinician deemed for any other reason that participation in the study might be
AN
contraindicated.
M

Having obtained informed written consent for participation and surgery, participant
D

characteristics at baseline were obtained, including: mean age; sex; ethnicity; highest
TE

education level attained; current smoking status; the short-form Oral Health Impact Profile

(OHIP-14)10; Locker’s global oral health item11; the Dental Anxiety Scale (DAS)12; history of
EP

third molar pain; the type and number of mandibular impactions; and, the number of

mandibular and maxillary third molars to be removed. Any active pericoronitis associated
C

with third molars intended for removal was treated and resolved prior to surgery.
AC

We assigned participants randomly to either the control group or the intervention group.

Block randomisation13 was the method used to generate the random allocation sequence. The

reason for this choice was to ensure balanced randomisation (that is, groups of approximately

equal number) right throughout the course of the study. The biostatistician involved in this

study generated this sequence. Implementation of the random allocation sequence was
 ACCEPTED
delegated to the hospital pharmacist, the only MANUSCRIPT
nonblinded person involved in the study. Based

off-site at the Dunedin Public Hospital, this individual was the only person in possession of the

random allocation sequence. Accordingly, as and when participants were recruited, the

pharmacist was asked to assign participants to their groups by simply progressing down the

allocation sequence. Each participant was given a randomisation number by the pharmacist.

PT
Participants allocated to the control group received paracetamol 1000mg plus a placebo orally

RI
6 hourly, and ibuprofen 400mg orally 8 hourly, commencing immediately postoperatively and

SC
continuing for at least the 48-hour duration of the study. The placebo was a gluten-free,

artificial sweetener containing aspartame, and it was similar to a codeine tablet in appearance.

U
Participants in the intervention group received the identical regimen to the control group
AN
except that the placebo had been replaced by codeine 60mg. All participants were operated on

by the same surgeon, who used a standardised surgical technique. All participants required
M

bone removal with a drill for at least one mandibular third molar. All surgical procedures were
D

carried out under aseptic conditions, intravenous sedation with midazolam (titrated dose) and
TE

dexamethasone 8mg, and with mepivacaine 2% (1:100 000 adrenaline) local anaesthetic for

blocks and infiltration. Additionally, all participants were asked to use 0.2% chlorhexidine
EP

mouthwash three times daily for one week, commencing the day after surgery.
C

If participants required additional postoperative pain relief, the oral and maxillofacial surgery
AC

team responsible for the study was available 24 hours a day via its on-call service through

Dunedin Public Hospital.

Outcome variables

Postoperative pain was primarily measured using a non-graduated 100mm visual analogue

scale (VAS), labelled “no pain” at the left pole and “worst pain imaginable” at the right pole.
 ACCEPTED
Each participant made a baseline MANUSCRIPT
postoperative VAS measurement, followed by multiple

postoperative VAS measurements recorded at home every 3 hours, while awake, for the first

48 hours following surgery.

Postoperative pain was also assessed using a postoperative questionnaire completed during a

PT
review appointment with the surgeon on the third day following surgery. This comprised a

series of items on the participant’s experience of postoperative pain, adherence to the

RI
prescribed treatment schedule, and whether rescue analgesia or other medical intervention

SC
had been necessary. One item exploring the participant’s overall perception of postoperative

pain used a verbal rating scale, with the response options of no pain, mild pain, moderate pain,

U
severe pain, or excruciating pain/agony.
AN
Data upon which to undertake a power analysis for the current study were scarce. However,
M

for an anticipated effect size of 0.5, an α value of 0.05 and a power of 0.8 to detect a difference,
D

the estimated number per group was 51.

TE

Statistical analyses
EP

The VAS data were analysed using an intention-to-treat approach because this gives a more

valid assessment of treatment efficacy in real-life clinical practice. Specifically, a modified

C

intention-to-treat approach was used because one participant was missing all VAS data
AC

subsequent to the baseline VAS and so was left out of the analysis. A linear mixed model (with

the individual as a random effect) was used.

Results

A total of 150 patients were assessed for eligibility to participate in the study (Figure 1). Of

these, 19 did not meet the eligibility criteria and were excluded. All remaining 131 patients
 ACCEPTED
(87.3%) consented to participate and so were MANUSCRIPT
enrolled in the study and subsequently allocated

randomly to either the control group or the intervention group. The random allocation process

yielded 67 participants in the control group, all of whom received their allocated control

regimen, and 64 participants in the intervention group, all of whom received their allocated

intervention. No participant in either group was lost to follow-up. However, 3 participants in

PT
the control group discontinued their prescribed treatment: 1 did not adhere to the treatment

schedule, and 2 required rescue medication. Similarly, 2 participants in the intervention group

RI
discontinued their prescribed treatment: 1 did not adhere to the treatment schedule, and 1

SC
required rescue medication. These participants were not excluded from the data analysis. All

67 participants in the control group were included for analysis, but only 63 of the 64

U
participants from the intervention group were included for analysis, since the excluded
AN
participant did not submit any VAS data at all.
M

Baseline data
D

The two groups were similar at baseline for age, sex, ethnicity, education level, smoking status,
TE

mean OHIP-14 score, impact prevalence, and DAS data (Table 1). They were also similar in

their third molar pain at baseline prior to surgery (Table 2), the type and number of
EP

mandibular third molar impactions, and the total number of mandibular and maxillary third

molars removed per participant (Table 3).

C
AC

Outcome and estimation

Figure 2 shows the mean VAS score at each time point for both groups. There does not appear

to be any apparent pattern of response (increasing/decreasing) or any sort of systematic

difference between the groups.

 ACCEPTED
Initially, a linear mixed model (with MANUSCRIPT
the individual as a random effect) was fitted, with the VAS

score as the dependent variable and the baseline VAS, intervention, time and an intervention-

by-time interaction term. Akaike’s Information Criterion (AIC) was used to determine whether

the interaction term was required in the model. The assumptions of the model were checked

and were found to be adequately met. The interaction term was found to be not required, and

PT
so it was omitted from the model.

RI
The linear mixed model showed that the baseline VAS score was associated with the

SC
subsequent VAS scores and that, as time passed, the VAS score increased (Table 4). With each

3hr period, the VAS score increased by an average of 0.08. The treatment effect was not

U
significant, indicating no difference in recorded pain levels between the control and the
AN
intervention group. The control group and the intervention group did not differ in their pain

over the first 48 hours following mandibular third molar surgery.

M
D

Analysis of the verbal rating scale for global postoperative pain, in which the five categories of
TE

this rating scale were treated as an ordinal variable, demonstrated no statistically significant

difference between the control and intervention groups in overall postoperative pain
EP

experienced over the first 48 hours following third molar surgery (Table 5). To confirm this,

the rating scale was additionally treated as a continuous variable and the means were
C

compared. Once again, there was no significant difference between the two groups (p=0.16).
AC

Adverse events

During the follow-up period, 12 of the 67 control participants (18%) and 4 of the 64

intervention participants (6%) claimed that their prescribed treatment schedule did not

provide sufficient pain relief for the first 48 hours following third molar surgery. Of these, 2

participants from the control group resorted to rescue medication, with both claiming that the
 ACCEPTED
alternative medication provided MANUSCRIPT
sufficient pain relief. Similarly, 1 participant from the

intervention group resorted to rescue medication but found no additional pain relief from the

alternative medication. Finally, no participant from the control group reported visiting a

medical practitioner for postoperative pain, whereas 2 participants from the intervention

group did so.

PT
Discussion

RI
This study confirmed the hypothesis that the addition of codeine 60mg 6 hourly to a regimen of

SC
paracetamol 1000mg 6 hourly and ibuprofen 400mg 8 hourly does not improve analgesia

following mandibular third molar surgery involving bone removal. The finding is consistent

U
with previous evidence showing that codeine-containing regimens are of no greater efficacy
AN
than other postoperative analgesia regimens for third molar surgery.1 5 14-21
M

Codeine is a weak opioid with questionable analgesic benefit. It is associated with potential
D

adverse effects including nausea, vomiting, constipation, sedation, dizziness, seizures, and
TE

respiratory depression. Moreover, it is an unpredictable prodrug because some people are

unable to convert codeine into morphine for effective analgesia.8 9 Codeine also has the
EP

potential for substance abuse and addiction in susceptible individuals.22 Opioid use, misuse

and addiction is on the rise, and opioid addiction has reached epidemic proportions in many
C

countries including the USA, Canada, various European nations, Australia, New Zealand, and
AC

South Asia.23 24 These issues ought to prompt a warning for clinicians on the prescription of

codeine for patients. There are safer, more effective analgesics available.

To our knowledge, this study is the first to evaluate pain relief following third molar surgery by

comparing paracetamol 1000mg/ibuprofen 400mg/placebo with paracetamol

1000mg/ibuprofen 400mg/codeine 60mg. The main strength of the study was that
 ACCEPTED
participants were both randomly allocatedMANUSCRIPT
and blinded for the intervention, and all

investigators were blinded. Additional strengths included its multi-dosing protocol, 48-hour

follow-up duration, multiple VAS recordings, and standardisation of treatment.

Missing VAS information was a limitation. Participant sleep periods are a plausible explanation

PT
for the majority of these missing data. A second (and arguably minor) explanation for some of

the missing VAS data could be unintentional non-adherence by participants. This issue may

RI
have been improved by a simpler study protocol for VAS data collection, such as standardising

SC
the time of surgery and postoperative VAS recordings. For example, all operations commence

at 9am with postoperative VAS recordings at 12pm, 3pm, 6pm, and 9pm on the day of surgery,

U
9am, 12pm, 3pm, 6pm, 9pm the following day, and 9am and 12pm the next day. Another
AN
limitation was that the sample mainly comprised young adults. This was unavoidable because

the greatest need for third molar surgery is among people in their late teenage years and early
M

twenties. A third limitation was the study duration. While its 48-hour duration is longer than
D

many comparable studies and hence may be considered a strength, 48 hours may equally be
TE

viewed as a limitation because pain after third molar surgery often continues after the second

day. Nevertheless, a 48-hour duration was decided on because it is clinically relevant, and a
EP

study of longer duration is likely to have resulted in poorer adherence to data collection.7 A

further limitation was omission of intraoperative complications (such as nerve injury or

C

displaced tooth) which may contribute to pain in the postoperative period. Finally, the study
AC

may have been more robust if medication side-effects, length of surgery, and potential

postoperative complications were also evaluated.

In conclusion, this study has shown that codeine does not provide better analgesia following

third molar surgery when added to a paracetamol 1000mg/ibuprofen 400mg regimen.

Dentists, oral surgeons, and oral and maxillofacial surgeons should be discouraged from using
 ACCEPTED
codeine after third molar surgery MANUSCRIPT
because there are safer, more effective alternatives for pain

relief. Such a recommendation may meet some opposition from clinicians who routinely

prescribe codeine and who have anecdotal evidence to the contrary. However, we assert a

movement away from codeine for postoperative pain management in third molar surgery is in

the patient’s best interest.

PT
Funding and treatment fees

RI
All consumable expenses (such as medications, surgical materials, cannulas), surgical

SC
equipment, sedation equipment, and University staff required for this study were funded

completely by the University of Otago. Patient payment for the surgical removal of their third

U
molars under intravenous sedation followed the standard fee policy set by the School of
AN
Dentistry, University of Otago.
M

Conflicts of interest
D

The authors declare that there were no conflicts of interest in relation to this study. No
TE

individual had a financial or personal relationship with any party or product that could bias

this research.
EP

Additional contributions
C

We thank Andria McGhie, the clinical trials pharmacist at Dunedin Public Hospital, for
AC

assigning participants to their groups and for preparing and dispensing the medications used

in this study. A word of appreciation is also reserved for all auxiliary staff involved in this

research (dental assistants, registered nurses, receptionists, and administrators) who work in

the Department of Oral Diagnostic and Surgical Sciences at the University of Otago.

References
 ACCEPTED
1 Ziccardi VB, Desjardins PJ, Daly-DejoyMANUSCRIPT
E, Seng GF: Single-dose vicoprofen compared

with acetaminophen with codeine and placebo in patients with acute postoperative pain

after third molar extractions. J Oral Maxillofac Surg 58:622-28, 2000.

2 Barden J, Edwards JE, McQuay J, Moore RA: Pain and analgesic response after third

molar extraction and other postsurgical pain. Pain 107:86-90, 2004.

PT
3 Haglund B, von Bülzingslöwen I: Combining paracetamol with a selective

cyclooxygenase-2 inhibitor for acute pain relief after third molar surgery: a randomized,

RI
double-blind, placebo-controlled study. Eur J Oral Sci 114:293-301, 2006.

SC
4 Hyllested M, Jones S, Pedersen JL, Kehlet H: Comparative effect of paracetamol, NSAIDs

or their combination in postoperative pain management: a qualitative review. Br J

U
Anaesth 88:199-214, 2002.
AN
5 Jung Y-S, Kim DK, Kim M-K, Kim H-J, Cha I-H, Lee E-W: Onset of analgesia and analgesic

efficacy of tramadol/acetaminophen and codeine/acetaminophen/ibuprofen in acute

M

postoperative pain: a single-center, single-dose, randomized, active-controlled, parallel-

D

group study in a dental surgery pain model. Clin Ther 26:1037-45, 2004.
TE

6 Van Dyke T, Litkowski LJ, Kiersch TA, Zarringhalam NM, Zheng H, Newman K:

Combination oxycodone 5mg/Ibuprofen 400mg for the treatment of postoperative pain:

EP

a double-blind, placebo- and active-controlled parallel-group study. Clin Ther 26:2003-

14, 2004.
C

7 Merry AF, Gibbs RD, Edwards J, Ting GS, Frampton C, Davies E, Anderson BJ: Combined
AC

acetaminophen and ibuprofen for pain relief after oral surgery in adults: a randomized

controlled trial. Br J Anaesth 104:80-8, 2010.

8 Phero JC, Becker D: Rational use of analgesic combinations. Dent Clin North Am 46:691-

705, 2002.

9 Stamer UM, Stuber F: Genetic factors in pain and its treatment. Curr Opin Anaesthesiol

20:478-84, 2007.
10 Slade GD: Derivation ACCEPTED
and validationMANUSCRIPT
of a short-form oral health impact profile.

Community Dent Oral Epidemiol 25:284-90, 1997.

11 Thomson WM, Mejia GC, Broadbent JM, Poulton R: Construct validity of Locker’s global

oral health item. J Dent Res 91:1038-42, 2012.

12 Corah NL: Development of a dental anxiety scale. J Dent Res 48:596, 1969.

PT
13 Roberts C, Torgerson D: Randomisation methods in controlled trials. BMJ 317:1301,

1998.

RI
14 Brown JD, Daniels SE, Bandy DP, Ko AT, Gammaitoni A, Mehta A, Boice JA, Losada MC,

SC
Peloso PM: Evaluation of multiday analgesia with etoricoxib in a double-blind,

randomized controlled trial using the postoperative third molar extraction dental pain

U
model. Clin J Pain 29:492-98, 2013.
AN
15 Daniels SE, Bandy DP, Christensen SE, Boice J, Losada MC, Liu H, Mehta A, Peloso PM:

Evaluation of the dose range of etoricoxib in an acute pain setting using the
M

postoperative dental pain model. Clin J Pain 27:1-8, 2011.

D

16 Daniels SE, Goulder MA, Aspley S, Reader S: A randomised, five-parallel-group, placebo-

TE

controlled trial comparing the efficacy and tolerability of analgesic combinations

including a novel single-tablet combination of ibuprofen/paracetamol for postoperative

EP

dental pain. Pain 152:632-42, 2011.

17 Joshi A, Parara E, Macfarlane TV: A double-blind randomized controlled clinical trial of

C

the effect of preoperative ibuprofen, diclofenac, paracetamol with codeine and placebo
AC

tablets for relief of postoperative pain after removal of impacted third molars. Br J Oral

Maxillofac Surg 42:299-306, 2004.

18 Lysell L, Anzen B: Pain control after third molar surgery: a comparative study of

ibuprofen (Ibumetin) and a paracetamol/codeine combination (Citodon). Swed Dent J

16:151-60, 1992.
 ACCEPTED
19 Forbes JA, Kehm CJ, Grodin MANUSCRIPT
CD, Beaver WT: Evaluation of ketorolac, ibuprofen,

acetaminophen, and an acetaminophen-codeine combination in postoperative oral

surgery pain. Pharmacother 10:94S-105S, 1990.

20 Habib S, Matthews RW, Scully C, Levers BGH, Shepherd JP: A study of the comparative

efficacy of four common analgesics in the control of postsurgical dental pain. Oral Surg

PT
Oral Med Oral Pathol 70:559-63, 1990.

21 Dionne RA, Campbell RA, Cooper SA, Hall DL, Buckingham B: Suppression of

RI
postoperative pain by preoperative administration of ibuprofen in comparison to

SC
placebo, acetaminophen, and acetaminophen plus codeine. J Clin Pharmacol 23:37-43,

1983.

U
22 Woolf AD, Greco C: Why can’t we retire codeine? Pediatr 133:e1354-55, 2014.
AN
23 Duncan, B: A rising tide. Of Substance: The National Magazine on Alcohol, Tobacco and

Other Drugs 10:20-3, 2012.

M

24 Erasmus J: The ugly side of codeine. New Zealand Dent Assoc News 162:45-6, 2012.
D
TE
C EP
AC
 ACCEPTED
Table 1: Participant characteristics at baseline,MANUSCRIPT
by group (brackets contain percentages unless
otherwise indicated).

Control (n=67) Intervention (n=64)

Mean age (SD) 24.1 (5.9) 23.0 (3.4)

Sex

PT
Number of males 22 (32.8) 25 (39.1)
Number of females 45 (67.2) 39 (60.9)

RI
Ethnicity
Pakeha 44 (65.7) 42 (65.6)
Other 23 (34.3) 22 (34.4)

SC
Highest education level attained
Secondary 15 (22.4) 14 (21.9)
Tertiary 52 (77.6) 50 (78.1)

Current smoking status

U
AN
Smoker 7 (10.4) 11 (17.2)
Non-smoker 60 (89.6) 53 (82.8)

OHIP-14a at baseline
M

Mean OHIP (SD) 12.7 (10.8) 10.9 (8.7)

Prevalence of 1+ impacts 31 (46.3) 22 (34.4)
D

DASb at baseline
Mean DAS (SD) 8.9 (3.4) 9.0 (3.5)c
TE

Dental anxiety
Dentally anxious 13 (19.4) 11 (17.5)c
Not dentally anxious 54 (80.6) 52 (82.5)
EP

aOHIP-14 = oral health impact profile – short version

bDAS = dental anxiety scale
cData missing for 1 individual
C
AC
 ACCEPTED
Table 2: Third molar pain at baseline, by groupMANUSCRIPT
(brackets contain percentages unless
otherwise indicated).

Control (n=67) Intervention (n=64)

History of third molar pain

Yes 56 (83.6) 56 (87.5)
No 11 (16.4) 8 (12.5)

PT
Third molar pain in last 4 weeks
Always 5 (7.5) 4 (6.3)

RI
Often 13 (19.4) 17 (26.6)
Sometimes 12 (17.9) 13 (20.3)
Occasionally 15 (22.4) 14 (21.9)
Never 22 (32.8) 16 (25.0)

SC
Mean preoperative VASa (SD) 3.6 (8.3) 6.3 (14.2)

U
aVAS = visual analogue scale (range of possible scores, 0-100mm)
AN
M
D
TE
C EP
AC
 ACCEPTED
Table 3: Type and number of mandibular MANUSCRIPT
impactions, and number of mandibular and
maxillary third molars removed, by group (brackets contain percentages).

Control (n= 67) Intervention (n= 64)

Mandibular impaction type

Mesioangular 38 (30.6) 42 (36.5)
Distoangular 25 (20.2) 31 (27.0)

PT
Vertical 30 (24.2) 21 (18.3)
Horizontal 30 (24.2) 21 (18.3)
Transverse 1 (0.8) 0 (0.0)

RI
Inverted 0 (0.0) 0 (0.0)

Number and type of third molars

removed

SC
1 mandibular only 5 (7.5) 5 (7.8)
1 mandibular + 1 maxillary 2 (3.0) 6 (9.4)
1 mandibular + 2 maxillary 3 (4.5) 2 (3.1)

U
2 mandibular only 12 (17.9) 16 (25.0)
2 mandibular + 1 maxillary 11 (16.4) 9 (14.1)
AN
2 mandibular + 2 maxillary 34 (50.7) 26 (40.6)
M
D
TE
C EP
AC
Table 4: Linear mixed model. ACCEPTED MANUSCRIPT

Outcome VASa Coefficient SEb z score P>z 95% CIc

Baseline VAS 0.571 0.118 4.830 0 0.339, 0.803

Time 0.076 0.028 2.750 0.006 0.022, 0.130
Intervention (B) -3.211 3.288 -0.980 0.329 -9.655, 3.233
Constant 20.778 2.492 8.340 0 15.894, 25.663

PT
aVAS = visual analogue scale (range of possible scores, 0-100mm)
bSE = standard error

RI
cCI = confidence interval

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED
Table 5: Global pain level over the MANUSCRIPT
first 48 hours following third molar surgery, by group
(brackets contain percentages).

Pain Level Control (n=67) Intervention (n=64) P

value
0.52
No pain 6 (9.0) 7 (10.9)
Mild 30 (44.8) 37 (57.8)
Moderate 25 (37.3) 16 (25.0)

PT
Severe 5 (7.5) 3 (4.7)
Excruciating (agony) 1 (1.5) 1 (1.6)

RI
U SC
AN
M
D
TE
C EP
AC
Figure Legend ACCEPTED MANUSCRIPT

Figure 1: CONSORT flow chart

Figure 2: Mean pain VAS over time, by group

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC