| |
 The
treatment of pain was a major interest of Beecher and Houde,
two of the early proponents of randomized clinical trial
methodology after its development in the late 1940’s.
These researchers called attention to challenges in analgesic
studies that differed from those encountered in the study
of structural diseases. These issues include the large magnitude
of placebo effects, the difficulty of symptom measurement,
and the frequent findings of negative results even with
drugs known to be effective analgesics. In a classic textbook
that discusses clinical trials totreat pain, insomnia, nausea,
and other symptoms, Beecher
(1959) pointed out that the same challenges beset studies
of diverse symptoms:
"The
measurement of pain has never seemed to this writer to be
only an end in itself, valuable as this end might be, but
rather an area where he could learn how to attack other subjective
responses."
Surprisingly,
this insight has been neglected. There has been rather little
interaction among investigators or regulators interested in
clinical trials in different symptoms, who generally associate
with others interested in the same group of diseases. I would
suggest that the accumulated literature of over 15,000 analgesic
clinical trials as well as writings on study design (Houde
et al., 1965; Houde
et al., 1966; Max
et al., 1991) and pain measurement (Price,
1988; Chapman
and Loeser, 1989; Turk
and Melzack, 1992) may offer useful lessons for the design
of studies of other symptoms. In this chapter, I will try
to point out some of the firm conclusions from the experience
of analgesic research. I will also highlight areas of analgesic
trial methodology that have had relatively little exploration--particularly
repeated dose studies in chronic pain--where an ongoing dialogue
among symptom researchers might move all of these fields ahead.
SOME
FUNDAMENTAL QUESTIONS ABOUT CLINICAL TRIALS OF TREATMENTS
FOR PAIN
Can
pain be measured scientifically?
Experienced
research physicians unacquainted with pain and symptom research
often ask me this, based on their observation that people
vary in their description of and emotional reaction to everyday
trauma, medical procedures, and chronic illness.
A brief
but convincing answer to this question is provided by Table
1.1 Even before most of the sophisticated work of modern
pain psychologists, clinical trials of opioid analgesics
in the 1950’s showed that using groups of 20-50 patients
with acute postoperative or cancer pain, different investigators
could readily replicate estimates of the relative potency
of different opioids to morphine, using simple 4- and 5-item
category scales for pain intensity
and relief.
| Table
1.1: Replicability of Single-Dose Analgesic Comparisons
Demonstrates the Validity of Simple Subjective Measures
of Pain |
| Drug
| Patient
population
| Investigator
| Potency
relative to morphine
|
|---|
| Oxymorphone | Cancer | Wallenstein
and Houde | 9.8 |
| Cancer | Eddy
and Lee | 9.9 |
| Postoperative | Dekornfeld | 10.0 |
| Phenazocine | Cancer | Houde
et al. | 3.2 |
| Postoperative | Dekornfeld
and Lasagna | 3.3 |
| Dextromoramide | Cancer | Bauer
et al. | 2.0 |
| Postoperative | Keats
et al. | 1.9 |
| Cancer | Houde
and Wallenstein | 1.3 |
| Dipipanone | Cancer | Houde
and Wallenstein | 0.49 |
| Cancer | Seed | 0.47 |
| Cancer | Cochin | 0.88 |
| Normorphine | Cancer | Houde
and Wallenstein | 0.40 |
| Postoperative | Lassagna | 0.25 |
| Dihydrocodeine | Cancer | Seed
et al. | 0.15 |
| Postoperative | Keats
et al. | 0.17 |
| Postoperative | Beecher | 0.17 |
| Relative
potency estimates of opioid analgesics to morphine.
Using simple subjective pain category and relief
scales, many independent groups obtained similar
results (From Houde et al., 1965). |
|
OPTIONAL
MATERIAL: BRIEF REVIEW OF PAIN MEASURES USED AS OUTCOMES
OF ANALGESIC CLINICAL TRIALS.
|
