BMJ 2001;322:1154
( 12 May )
Laurie Allan
a Chronic Pain Services, Northwick
Park and St Mark's NHS Trust, Harrow, Middlesex HA1 3UJ, b Department
of Family Medicine, University of Alberta, Edmonton, Alberta, Canada T6G 2C8, c Multidisciplinary
Pain Centre, Department of Anaesthesiology, Herlev University Hospital,
DK-2730, Denmark, d Cliniques Universitaires St-Luc,
1200 Brussels, Belgium, e Alg Ziekenhuis Eemland De
Lichtenberg, 3818 ES Amersfoort, Netherlands, f Strand
Private Hospital, Cape Town 7139, South Africa, g Helsinki
University Central Hospital Pain Clinic, 00290 Helsinki, Finland
Correspondence to: L Allan [log in to unmask]
|
Abstract |
Objectives: To compare patients' preference for
transdermal fentanyl or sustained release oral morphine, their level
of pain control, and their quality of life after treatment.
Design: Randomised, multicentre,
international, open label, crossover trial.
Setting: 35 centres in Belgium,
Canada, Denmark, Finland, the United Kingdom, the Netherlands, and
South Africa.
Participants: 256 patients (aged
26-82 years) with chronic non-cancer pain who had been treated
with opioids.
Main outcome measures: Patients'
preference for transdermal fentanyl or sustained release oral
morphine, pain control, quality of life, and safety assessments.
Results: Of 212 patients,
138 (65%) preferred transdermal fentanyl, whereas 59 (28%)
preferred sustained release oral morphine and 15 (7%) expressed
no preference. Better pain relief was the main reason for preference
for fentanyl given by 35% of patients. More patients considered pain
control as being "good" or "very good" with
fentanyl than with morphine (35% v
23%, P=0.002). These results were reflected in both patients' and
investigators' opinions on the global efficacy of transdermal
fentanyl. Patients receiving fentanyl had on average higher quality
of life scores than those receiving morphine. The incidence of
adverse events was similar in both treatment groups; however, more
patients experienced constipation with morphine than with fentanyl
(48% v 29%, P<0.001). Overall,
41% of patients experienced mild or moderate cutaneous problems associated
with wearing the transdermal fentanyl patch, and more patients
withdrew because of adverse events during treatment with fentanyl
than with morphine (10% v 5%).
However, within the subgroup of patients naive to both fentanyl and
morphine, similar numbers of patients withdrew owing to adverse
effects (11% v 10%, respectively).
Conclusion: Transdermal fentanyl
was preferred to sustained release oral morphine by patients with
chronic non-cancer pain previously treated with opioids. The main
reason for preference was better pain relief, achieved with less
constipation and an enhanced quality of life.
What is already known on this topic Studies with transdermal fentanyl have
shown efficacy and preference over sustained release oral morphine in the
treatment of cancer pain What this study adds Both transdermal fentanyl and sustained
release oral morphine provided effective and well tolerated pain relief During fentanyl treatment patients
experienced superior pain relief, higher quality of life, and less
constipation; fentanyl was preferred to morphine by 65% of patients |
|
Introduction |
In 1998 the World Health Organization survey of nearly
26 000 patients from five continents reported that 22% had had persistent
pain sometime over the previous year.1 Pain is
one of the commonest reasons for visiting a doctor and, when pain is
chronic, the multimillion healthcare costs are only exceeded by the
multibillion costs of work loss, disability, and social welfare
benefits. Global socioeconomic costs at one end of the spectrum are
matched by personal suffering at the other. Pain is often
undertreated or mistreated, with patients going from doctor to
doctor for relief and finally moving outside mainstream medicine in
increasing numbers.2
Opioids are the mainstay of management of cancer pain, providing
effective pain relief. 3
4 Opioids
are the most powerful analgesics, but politics, prejudice, and
continuing ignorance still impede optimum prescribing.5 A review
of retrospective and survey data confirms the efficacy of opioids in
the treatment of chronic non-cancer pain and found that fears of
addiction were not justified.6 Randomised
controlled trials of intravenous opioids in chronic non-cancer pain
show benefit over placebo for morphine and fentanyl, whereas oral
placebo controlled trials show efficacy for codeine, morphine, and
oxycodone.7-11
Worldwide, the value of opioids in this role has led to the
development of management guidelines, with recommendations from national
organisations.12-15
Morphine is the standard opioid against which others are judged
and is usually prescribed in a sustained release oral formulation for
the treatment of chronic pain. 5 10 Severe
constipation, a persistent complication of oral opioids, may affect
some patients' quality of life more than their pain.16
Fentanyl, a lipid soluble synthetic opioid, can be delivered in a
transdermal controlled release formulation, providing continuous, controlled
systemic delivery of fentanyl for up to 72 hours.17 Studies
with transdermal fentanyl have shown analgesic efficacy in cancer
pain.18-22
Most patients preferred fentanyl, which was associated with less
constipation than morphine. 21 22 Also,
fentanyl has been shown to relieve neuropathic pain that is relatively
insensitive to opioids. 8 23
Opioids are individually titrated to an effective dose. Therefore
little difference would be expected between opioids in efficacy or
improvements in quality of life, which is confirmed by studies in
cancer pain. 5
6 24 25
Recognising the increasing importance of patients' preference and
choice, we investigated in a large, multicentre, two way crossover
trial whether patients with chronic non-cancer pain accustomed to
opioids would prefer transdermal fentanyl over sustained release
oral morphine, has been found in patients with cancer pain.21 We also
assessed pain control, quality of life, and adverse events.
|
Participants
and methods |
Protocol
The study was approved by local ethics committees, and patients
giving written, informed consent were recruited from 35 specialist
pain clinics in Belgium (15 patients), Canada (58), Denmark
(67), Finland (16), the United Kingdom (65), the Netherlands (21),
and South Africa (14).
Patients were invited to participate if
they were aged over 18 years and had chronic non-cancer pain requiring
continuous treatment with potent opioids for six weeks preceding the
trial. They had to have achieved moderate pain control with a stable
dose of oral opioid for seven days before the trial. Exclusion criteria
included pain not responding to opioids, a history of allergy or
hypersensitivity to opioids, life threatening disease, skin disease
precluding the transdermal system, reduced level of consciousness,
social isolation, concomitant psychiatric disorders, history of
substance misuse, clinically relevant cardiac, nervous system, or
respiratory disease, participation in another clinical research
project, and possible pregnancy or lactation.
Treatments
Baseline assessment included recording the patients'
characteristics, medical history, physical abnormalities, and vital
signs. Efficacy and safety data were collected at day 7, 16, and
28 of each treatment period. Patients' pain was classified according
to the International Association for the Study of Pain as
nociceptive (stimulation of intact nociceptors by noxious stimuli), neuropathic
(disease or trauma of the nervous system), or combined nociceptive
and neuropathic (table 1).
|
The requirement for opioid was determined
individually over the 24 hours before the first dose of study drug was
given. Analgesic doses of fentanyl equivalent to the patients'
previous opioid dose were calculated from the manufacturer's
recommendations.26
Patients were treated with fentanyl patches (Durogesic, Janssen-Cilag)
releasing 25, 50, 75, or 100 µg fentanyl/hr and
sustained release oral morphine as
10, 30, 60, 100, or 200 mg tablets (MS Contin, Napp
Laboratories). Patients receiving morphine received half the
equivalent analgesic dose every 12 hours. At crossover, patients received
the same opioid dose as before the study. Patients randomised to
receive fentanyl first were given the first dose of morphine when
the last fentanyl patch was removed. Patients randomised to receive
morphine first had the first fentanyl patch applied when the last
dose of morphine was given.
Patients were prescribed immediate release
morphine (initially 5 mg) every four hours as needed. Patients requiring
more than 60 mg of this rescue drug over two days of a three
day period with fentanyl could increase their fentanyl dose.
Patients receiving morphine needing more than two doses of the
rescue drug per day could titrate to a higher dose of morphine.
Subject preferencesThe primary efficacy variable was the patient's preference for
transdermal fentanyl or sustained release oral morphine and their
main reason for preference. This evaluation was completed either at
the end of the trial or at the end of treatment in patients who
withdrew before completion.
Pain control and
treatment assessmentAt each visit
patients were assessed for pain control compared with the previous visit. Both
investigator and patient completed a global treatment assessment at
the end of each treatment period.
Rescue drugPatients recorded their use of rescue morphine for breakthrough or
incident pain. The use of the drug in the first week of each study
period was excluded from the analysis, which was considered the
"titration period."
Quality of life
assessmentQuality of life
(SF-36) and pain intensity (0 being low and 100 high) were assessed
at baseline and at the end of each treatment period.27
Safety observationsThe safety evaluation at the first visit, at crossover, and at the
end of the trial included a physical examination, standardised
measurement of vital signs, and overall assessment of disease
progression. Details of all adverse events and presumed relation to
the drugs were noted by the investigator. At each visit bowel
function was assessed by using a questionnaire, and the application
site of the fentanyl patch was evaluated.
Statistical analysesThe primary efficacy variable was analysed with a binomial test
determining whether the proportion of patients who either
"preferred" or "very much preferred" a treatment was
larger than 0.5. Differences in personal variables at baseline between
treatment groups were analysed with the Van Elteren test for
continuous variables and the Cochran-Mantel Haenszel test for
categorical variables, both adjusted for country. The mean daily
dose of rescue drug, assessment of global treatment, assessment of
pain control, and quality of life scores were compared with the Koch
non-parametric paired analysis for crossover designs and adjusted
for country.28
Order effects were assessed by adding an interaction variable for
treatment sequence and were excluded if P>0.10. All other P
values less than 0.05 were considered significant.
AssignmentPatients were assigned to treatment groups by using the central
randomisation minimisation technique.29 One
group was randomised to four weeks of treatment with sustained
release oral morphine followed by transdermal fentanyl for four
weeks. The second group received the same treatments but in reverse
order. A washout period was considered unethical.
|
Results |
Participant flow
The figure shows the progress of patients through the study. Sixty
patients withdrew; 37 because of adverse events, five because
of insufficient efficacy, and 18 for other reasons. Five
patients without baseline data were excluded from the efficacy
analysis. All patients were included in the safety analysis.
|
Baseline characteristics,
including type of pain and patients' previous use of opioids, did not differ
between the two groups (tables 1 and 2). The mean
starting dose of transdermal fentanyl was 39.7 µg/hr (range
25-200 µg/hr) and of sustained release oral morphine
123.0 mg/24 hr (range 10-700 mg/24 hr). The mean dose of
fentanyl at the end of the study was 57.3 µg/hr (range 0-325 µg/hr)
and of morphine 133.1 mg/24 hrs (range 0-800 mg/24 hrs).
|
Analysis
Patient preference
Preference could not be assessed in 39 of 251 patients,
leaving a total of 212 patients for analysis. A higher
proportion of patients preferred or very much preferred transdermal
fentanyl to oral sustained release morphine (138 (65%) v 59 (28%); P<0.001) (table 3). Fifteen patients
(7%) did not express a preference. After exclusion of
24 patients with a "bad" or "very bad"
score while taking morphine before the study, 69% of patients
expressed a "strong" or "very strong" preference for fentanyl.
Patient preference for fentanyl was not
significantly different in patients with nociceptive pain (75 of
108, 69%), neuropathic pain (31 of 53, 58%), or mixed
neuropathic and nociceptive pain (32 of 51, 63%). The
predominant reason given for preferring fentanyl was better pain
relief, followed by greater convenience and fewer adverse events
(table 3). In a
subgroup of 66 patients who were neither accustomed to fentanyl
nor morphine (they had taken other opioids before the study), 62%
preferred fentanyl.
Pain control, treatment assessments, and
rescue drug
Patients treated with transdermal fentanyl had on average lower pain
intensity scores than those treated with sustained release oral
morphine (mean 57.8, range 33.1-82.5 v mean 62.9, range 41.2-84.6;
P<0.001), irrespective of the order of treatment. More patients
receiving fentanyl considered their pain control to be good or very
good than those receiving morphine (35% v 23%, P=0.002) (table 4). Similar
satisfaction was found among patients receiving fentanyl with
nociceptive pain (43 of 123, 35%), neuropathic pain
(21 of 62, 34%), or combined neuropathic and nociceptive
pain (23 of 63, 37%). The corresponding satisfaction rates
with morphine were 15 or 59 (25%) patients, 27 of
116 (23%), and 12 of 59 (20%), respectively. Overall,
about one quarter of patients considered their pain control poor or
very poor with either treatment (table 4).
|
|
In the investigators' opinion, global
efficacy of fentanyl was good or very good in 131 of 225 (58%)
patients compared with 75 of 224 (33%) patients receiving
morphine (P<0.001). The corresponding percentages from the
patient assessments were 60% for fentanyl and 36% for morphine
(P<0.001).
Analysis of the consumption of rescue drug
during the last three weeks of each treatment period showed that the mean
(standard deviation) consumption was significantly higher with
fentanyl (29.4 (33.0) mg) than with morphine (23.6 (32.0)
mg; P<0.001). A significant (P<0.05) period effect was also
observed: the higher consumption during fentanyl treatment was more
apparent in the second trial period (mean 32.4 (SD 38.5) mg)
than the first (26.3 (26.0) mg), where the consumption of the
rescue drug remained essentially the same over the two treatment
periods in the morphine group (23.7 (35.3) mg v 23.6 (27.3) mg).
Quality of lifePatients had, on average, quality of life scores below the median
on a scale ranging from 0 to 100 (lowest to highest
wellbeing) (table 5).
Patients receiving transdermal fentanyl had higher overall quality
of life scores than patients receiving sustained release oral
morphine in each of eight categories measured by the SF-36.
Differences were significant in the categories for bodily pain,
vitality, social functioning, and mental health. No period effect was
noted.
|
Adverse effectsThe overall incidence of treatment related adverse events was
similar in both groups (table 6), as was the proportion
of patients with adverse events (74% v
70%). Transdermal fentanyl was associated with a higher incidence of
nausea (26% v 18%)
than was sustained release oral morphine, whereas constipation was
less common with fentanyl than with morphine (16% v 22%). Reduced constipation was confirmed by the
bowel function questionnaire (29% fentanyl v 48% morphine; P<0.001) (table 6). Erythema and itching
at application sites commonly occurred in patients receiving fentanyl
(101 of 250, 41%) but were of mild to moderate intensity. Few
patients had serious adverse events (2.8% v
3.8% for fentanyl and morphine, respectively), and only one patient,
in the morphine group, hypoventilated. No deaths occurred, and no
clinically important changes of vital signs were observed.
|
Patient withdrawalsWithin the total patient population the number of withdrawals in
the fentanyl group was almost double (16%) that in the morphine
group (9%) (table 7).
More patients withdrew because of adverse events during treatment
with transdermal fentanyl (11%) than with sustained release oral
morphine (4%). However, subgroup analysis of 66 patients who
had taken neither fentanyl nor morphine before the study showed that
both the total number of withdrawals and withdrawals for adverse
events were similar between treatment groups (table 7). Ten patients (16%)
in the fentanyl group and 11 patients (18%) in the morphine group
withdrew from the trial; seven patients receiving fentanyl (11%) and
six receiving morphine (9.8%) withdrew because of an adverse event.
|
|
Discussion |
Patients with chronic non-cancer pain generally preferred
treatment with transdermal fentanyl (65%) than with sustained release oral
morphine (28%). A similar result was observed in patients with
cancer pain. 21
22
Furthermore, our findings confirm that potent opioids can provide
satisfactory pain relief for the difficult clinical problem of
chronic non-cancer pain. Although recruitment bias cannot be
excluded, it cannot entirely explain the observed difference in
treatment outcome.
Despite preference and better pain relief, more patients withdrew
because of adverse events in the first fentanyl period than in the
first morphine period. The phenomenon of preference for an opioid
despite higher reporting of adverse events is well recognised in
blinded controlled trials. 7 9-11 Most
patients (76%) had taken morphine for six weeks before entry to the
study and would be accustomed to its side effects, making it
unlikely that they would report additional adverse events when
randomised to sustained release oral morphine. This may represent
"incomplete cross tolerance" leading to a greater than
anticipated potency. 3
30 It is
supported by analysis of a subgroup of patients who had taken
neither morphine nor fentanyl before the trial. Here, withdrawals in
total and in relation to adverse events were similar in both groups.
Comparisons of opioid action must be made at equianalgesic doses.
It would be possible to explain the observed improvement in pain
control and constipation with fentanyl if the initial dose ratios
were wrong. The dosage of fentanyl increased consistently during
each four week period, and these patients consumed more rescue drug
than those receiving morphine. These findings confirm that a higher
ratio of starting dose may be required compared with the
conservative equianalgesic dose table used in this study.19 However,
individual dose titration is vital and allows for the variability in
patients' response to different opioids and the reported need to
reduce the dose during "opioid rotation" in patients showing
toxicity. 3
21 31
The tables for equianalgesic dose derive from studies of single
doses in selected populations and should be regarded as tentative for
incomplete cross tolerance. 3 30 Our
patients were conditioned to opioids, mainly morphine, and switching
to fentanyl may partly explain the improved pain control. The switch
may have raised expectation of increased pain relief, partly
attributable to a placebo analgesic effect.32 Most
patients, however, preferred fentanyl regardless of the order of
treatment. Exclusion of patients dissatisfied with morphine did not
affect the percentage of patients preferring fentanyl.
The higher consumption of rescue drug during treatment with
fentanyl was small (5.8 mg/24 hr overall), and probably not clinically
important, but may reflect a less flexible dose titration with fentanyl.
Furthermore, the difference in consumption of rescue drug was not
significantly different between treatments in the first period.
Differences in pain relief may also be explained by selectivity of
opioid receptors. Indeed, recent research indicates a genetic basis
for differences in pain sensitivity and response to analgesics.33
A significantly lower incidence of constipation was detected in
the formal assessment of bowel function by patients receiving fentanyl
(29% fentanyl v 48% morphine;
P<0.001), confirming previous reports.18-22 In
rats, fentanyl has a more favourable dose-analgesia to
dose-constipation ratio than morphine, probably because the higher
lipid solubility of fentanyl enables it to pass through the
blood-brain barrier more easily than morphine.34 Giving
fentanyl transdermally limits gastrointestinal concentration compared
with oral morphine and consequently has less effect on opioid receptors
in the gut.
The fentanyl patch formulation affords a convenient system of
delivery over 72 hours. It may prevent "clock watching" and
breakthrough pain associated with shorter acting formulations, thus
improving compliance.4
In a "double dummy" design, preference for one delivery
system would have been difficult to assess if patients were
receiving both drugs together, particularly considering the
difficulties and risks associated with simultaneously titrating morphine
and fentanyl, as they have different dose schedules. Placebo effects
can explain analgesia but not poor analgesia.35 Therefore,
although a placebo effect is a possible explanation for our
findings, given an overtly different administration, it is a less
plausible explanation for those receiving fentanyl or morphine who
had poor pain control. These findings are consistent with other
reports that opioids do not provide adequate pain control to all
patients with chronic non-cancer pain. 5 9 10
Finally, we believe that using a pragmatic, clinical practice
based approach, particularly in a large sample size, is justified, especially
in the light of recent problems applying quality designs to clinical
trials. 36
37 The
"explanatory" (evidence based) approach requires a placebo
for comparison, whereas the "pragmatic" approach generally
compares a new treatment with the best in clinical use for the
particular clinical circumstances of patients.38 The
existence of a gold standard treatment allows direct comparison rather
than a placebo control, so that transdermal fentanyl can be directly
compared with sustained release oral morphine.39
Strong treatment preferences can present difficulties but may be
avoided by the crossover design. 40 41
Patients' preference, although important for all clinical decisions,
deserves special emphasis when diseases or treatments affect quality
of life, the treatment involves risks or side effects, or the choice
between treatments is a "close call."42 The
patient may be the best judge of the delicate balance between
analgesic efficacy, side effects, and the overall experience of
pain. This reflects our choice of patients' preference as the
primary efficacy variable. Furthermore, pragmatic outcome measures
such as quality of life and patients' preference may, ultimately,
form a more accurate evaluation of treatment effects than pain
measures alone.
|
Acknowledgments |
We thank all the investigators who
participated in the trial: J Maeyaert, L Plaghki (Belgium); J Clark, A Mailis,
D Moulin, M Ong-Lam, D Reid, P Watson (Canada); S Andersen, C
Christiansen, S Clemensen, K Glahn, T Jonsson, S Larsen, F Molke
Borgbjerg, A Schou Olesen, J Mųlgaard (Denmark); V Järvimäki, T
Heiskanen (Finland); R Atkinson, P Brown, F Campbell, R Gautam, M
Hanna, D Hughes, C Knight, W Notcutt (United Kingdom); G Braak, J
Helmers, G Van Oss, W Zuurmond (Netherlands); D Lines (South Africa).
Contributors: LA was the principal author of the paper. H Noorduin
(international supervisor of the trial), L Bijnens (biostatistics), L Haazen, M
Travers, D Peelmans, N Currie, A Jepsen, M Jarvinen, M Uitendaal, P Matthysen
(local trial coordination and monitoring). MT will act as guarantor for the
paper.
|
Footnotes |
Funding: The study was supported by a grant from Janssen Research
Foundation, Belgium.
Competing interests: LA receives support from both Janssen-Cilag,
the manufacturer of transdermal fentanyl (Durogesic) and Napp
Laboratories, the manufacturer of sustained release morphine. EK has
been reimbursed by Janssen-Cilag for participation at a meeting
sponsored by Janssen-Cilag.
|
References |
1. |
Gureje O, Von Korff M, Simon G, Gater R.
Persistent pain and well-being: A World Health Organization study in primary
care. JAMA 1998; 280: 147-151 |
2. |
Ashburn MA, Staats PS. Management of
chronic pain. Lancet 1999; 353:
1865-1869 |
3. |
Portenoy RK. Opioid and adjuvant
analgesics. In: Mitchell M, ed. Pain 1999an updated
review. Seattle: IASP Press, 1999:3-18. |
4. |
World Health Organization. Cancer pain relief, 2nd ed. Geneva: WHO,
1996. |
5. |
McQuay H. Opioids in pain management. Lancet 1999; 353: 2229-2232 |
6. |
Portenoy RK. Opioid therapy for chronic
non-malignant pain: a review of critical issues. J Pain Symptom Manage 1996; 11: 203-217 |
7. |
Rowbotham MC, Reisner-Keller LA, Fields
HL. Both intravenous lidocaine and morphine reduce the pain of postherpetic
neuralgia. Neurology 1991; 41:
1024-1028 |
8. |
Dellemijn PLI, Vanneste JAL. Randomised
double-blind active-placebo-controlled crossover trial of intravenous
fentanyl in neuropathic pain. Lancet
1997; 349: 753-758 |
9. |
Arkinstall W, Sandler A, Goughnour B,
Babul N, Harsanyi Z, Drake A. Efficacy of controlled-release codeine in
chronic non-malignant pain: a randomized, placebo-controlled clinical trial. Pain 1995; 62: 169-178 |
10. |
Moulin DE, Iezzi A, Amireh R, Sharpe
WKJ, Boyd D, Merskey H. Randomised trial of oral morphine for chronic
non-cancer pain. Lancet 1996;
347: 143-147 |
11. |
Watson CPN, Babul N. Efficacy of
oxycodone in neuropathic pain: a randomized trial in postherpetic neuralgia. Neurology 1998; 50: 1837-1841 |
12. |
Schug SA, Merry AF, Acland RH. Treatment
principles for the use of opioids in pain of nonmalignant origin. Drugs 1991; 42: 228-239 |
13. |
Brown RL, Fleming MF, Patterson JJ.
Chronic opioid analgesic therapy for chronic low back pain. J Am Board Fam Pract 1996; 9: 191-204 |
14. |
Graziotti PJ, Goucke CR. The use of oral
opioids in patients with chronic non-cancer pain. Management strategies. Med J Austr 1997; 167: 30-34 |
15. |
Anon. The use of opioids for the
treatment of chronic pain: a consensus statement from the American Academy of
Pain Medicine and the American Pain Society. Pain
Forum 1997; 6: 77-79 |
16. |
Cummings-Ajemian I. Treatment of related
symptoms. In: Patt RB, ed. Cancer pain,
section III, non-pharmacological treatment and novel approaches to management.
Philadelphia: JB Lippincott, 1993. |
17. |
Jeal W, Benfield P. Transdermal fentanyla review of its
pharmacological properties and therapeutic efficacy in pain control. Drugs 1997; 53: 109-138 |
18. |
Grond S, Zech D, Lehman KA, Radbruch L,
Breintenbach H, Hertel D. Transdermal fentanyl in the long-term treatment of
cancer pain: a prospective study of 50 patients with advanced cancer of
the gastrointestinal tract or the head and neck origin. Pain 1997; 69: 191-198 |
19. |
Donner B, Zenz M, Tryba M, Strumpf M.
Direct conversion from oral morphine to transdermal fentanyl: a multicentre
study in patients with cancer pain. Pain
1996; 64: 527-534 |
20. |
Donner B, Zenz M, Strumpf M, Raber M.
Long-term treatment of cancer pain with transdermal fentanyl. J Pain Symptom Manage 1998; 15: 168-175 |
21. |
Ahmedzai S, Brooks D. Transdermal
fentanyl versus sustained release oral morphine in cancer pain: preference,
efficacy, and quality of life. J Pain
Symptom Manage 1997; 13: 254-261 |
22. |
Payne R, Mathias SD, Pasta DJ, Wanke LA,
Williams R, Mahmoud R. Quality of life and cancer pain: satisfaction and side
effects with transdermal fentanyl versus oral morphine. J Clin Oncol 1998; 16: 1588-1593 |
23. |
Dellemijn PLI, van Duijn H, Vanneste
JAL. Prolonged treatment with transdermal fentanyl in neuropathic pain. J Pain Symptom Manage 1998; 16: 220-229 |
24. |
TTS-Fentanyl Multicentre Study Group.
Transdermal fentanyl in cancer pain. J
Drug Dev 1994; 6: 93-97 |
25. |
Zech DFJ, Grond SAU, Dauer HG,
Stollenwerk B, Lehmann KA. Transdermal fentanyl and initial dose-finding with
patient-controlled analgesia in cancer pain. A pilot study with
20 terminally ill cancer patients. Pain
1992; 50: 293-301 |
26. |
Southam MA. Transdermal fentanyl
therapy: system design, pharmacokinetics and efficacy. Anticancer Drugs 1995; 6(suppl 3): 29-34 |
27. |
Ware JE, Sherbourne CD. The MOS 36-item
short-form health survey (SF-36). I. Conceptional framework and item
selection. Med Care 1992; 30:
473-483 |
28. |
Koch CG. The use of non-parametric
methods in the statistical analysis of the two period change-over design. Biometrics 1972; 28: 577-584 |
29. |
Watson HR, Pearce AC. Treatment
allocation in clinical trials: randomisation and minimisation compared in
three test cases. Pharmaceut Med
1990; 4: 207-212 |
30. |
Collett B-J. Opioid tolerance: the
clinical perspective. Br J Anaesth
1998; 81: 58-68 |
31. |
Bruera E, Sloan P, Mount B, Scott J,
Suarez-Almazar M. A randomized, double-blind, double-dummy, crossover trial
comparing the safety and efficacy of oral sustained-release hydromorphine
with immediate-release hydromorphine in patients with cancer pain. J Clin Oncol 1996; 14: 1713-1717 |
32. |
Amanzio M, Benedetti F.
Neuropharmacological dissection of placebo analgesia: expectation-activated
opioid systems versus conditioning-activated specific subsystems. J Neurosci 1999; 19: 484-494 |
33. |
Uhl GR, Sora I, Wang Z. The mu opiate
receptor as a candidate gene for pain: polymorphisms, variations in
expression, nociception, and opiate responses. Proc NatlAcad Sci 1999; 96: 7752-7755 |
34. |
Megens AAHP, Artois K, Vermeire J, Meert
T, Awouters FHL. Comparison of the analgesic and intestinal effects of
fentanyl and morphine in rats. J Pain
Symptom Manage 1998; 15: 253-258 |
35. |
McQuay HJ, Jadad AR, Carroll D, Faura C,
Glynn CJ, Moore RA, et al. Opioid sensitivity of chronic pain: a
patient-controlled analgesia method. Anaesthesia
1992; 47: 757-767 |
36. |
Moore RA, Gavaghan D, Tramer MR, Collins
SL, McQuay HJ. Size is everythinglarge amounts of information are needed to overcome random
effects in estimating direction and magnitude of treatment effects. Pain 1998; 78: 209-216 |
37. |
Haynes B. Can it work? Does it work? Is
it worth it? BMJ 1999; 319:
652-653 |
38. |
Max M. Methodological issues in the
design of analgesic clinical trials. In: Mitchell M, ed. Pain 1999an update review. Seattle: IASP Press, 1999. |
39. |
Tramer MR. When placebo controlled
trials are essential and equivalence trials are inadequate. BMJ 1998; 317: 875-880 |
40. |
Torgerson DJ, Sibbald B. Understanding
controlled trials: what is a patient preference trial? BMJ 1998; 316: 360 |
41. |
Senn S. Within-patient studies:
cross-over trials and n-of-1 studies. In: Mitchell M, ed. Pain 1999an update review. Seattle: IASP Press, 1999. |
42. |
Goodare H, Lockwood S. Involving
patients in clinical research. BMJ
1999; 319: 724-725 |
(Accepted 13 December 2000)
Edward E.
Rylander, M.D.
Diplomat American
Board of Family Practice.
Diplomat American
Board of Palliative Medicine.