| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Health Quest Therapy and Research Institute, Austin, Texas; the Procter and Gamble Company, Health Care Research Center, Mason, Ohio; and the Department of Obstetrics and Gynecology, the University of Missouri-Kansas City, Truman Medical Center, Kansas City, Missouri.
Address reprint requests to: Roger P. Smith, MD, Department of Obstetrics and Gynecology, Truman Medical Center, 2301 Holmes Street, Kansas City, MO 64108, E-mail: BGumAlley{at}earthlink.net
 |
Abstract |
|---|
Top
 Abstract Materials and MethodsResultsDiscussionReferences |
|---|
Methods: We conducted a randomized placebo and active controlled (double dummy), parallel study using an abdominal patch (heated or unheated) for approximately 12 consecutive hours per day and oral medication (placebo or ibuprofen 400 mg) three times daily, approximately 6 hours apart for 2 consecutive days. Pain relief and pain intensity were recorded at 17 time points. There was at least 85% power to detect a true one-unit difference in the 2-day pain relief treatment means for comparisons with the unheated patch plus oral placebo group using a one-tailed test at the .05 level of significance, based on an observed within-group standard deviation of 1.147.
Results: Eighty-four patients were enrolled and 81 completed the study protocol. Over the 2 days of treatment, the heated patch plus placebo tablet group (mean 3.27, P < .001), the unheated patch plus ibuprofen group (mean 3.07, P = .001), and the combination heated patch plus ibuprofen group (mean 3.55, P < .001) had significantly greater pain relief than the unheated patch plus placebo group (mean 1.95). Greater pain relief was not observed for the combination heated patch plus ibuprofen group compared with the unheated patch plus ibuprofen group (P = .096); however, the time to noticeable pain relief was statistically significantly shorter for the heated patch plus ibuprofen group (median 1.5 hours) compared with the unheated patch plus ibuprofen group (median 2.79 hours, P = .01).
Conclusion: Continuous low-level topical heat therapy was as effective as ibuprofen for the treatment of dysmenorrhea.
The use of heating pads and hot-water bottles for the treatment of menstrual pain has a long history in common use and folklore.1 Unfortunately, this therapy lacks rigorous or systematic evaluation in the scientific literature,2 and only a few case reports exist.3 The recent development of small wearable devices capable of supplying a low level of topical heat at a constant temperature for a prolonged time now makes this modality portable enough to be a viable treatment option. This study was conducted to assess whether topically applied heat could provide subjective relief of the pain associated with menstruation.
|
Materials and Methods |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Eligibility was determined during two initial visits, the first to sign informed consent, complete a medical and dysmenorrhea history, and review the subject diaries; the second to have a brief physical examination including the measurement of abdominal skin temperature (by infrared thermometry), a pelvic examination, and Papanicolaou smear (if one had not been done in the preceding year) to confirm eligibility. A third visit was carried out to measure applied skin temperature, and a fourth (exit) visit was done within 3 to 7 days to finalize the study.
The study was a randomized, placebo and active controlled (double dummy), parallel study. Double dummy (placebo) design is commonly used when different treatment types are compared. Each treatment kit consisted of a device and a tablet. The intent of the double dummy design was to minimize potential bias associated with different treatment types.4
Women were randomly assigned to one of four treatment groups. Subjects were asked to wear a kidney bean–shaped ultra-thin medical device (heated or unheated) that adhered to the inside of the underwear on the lower abdominal region for approximately 12 consecutive hours per day for 2 consecutive days. In addition, subjects had to take two tablets of oral medication (matched placebo or ibuprofen 200 mg, for a total dose of 400 mg) three times daily, approximately 6 hours apart. The heat patch supplied heat at a constant temperature of approximately 38.9C over a surface area of 180 cm2. This device was capable of supplying warmth for the 12-hour study period. The position of the abdominal patch on the lower abdomen was standardized and the patch was held in place by adherence to the woman’s underwear, which completely covered the patch while in position.
The randomization and masking was the responsibility of the sponsor’s Clinical Supplies and Logistics department. A computerized randomization sequence generated the randomization schedule in blocks of four to ensure balanced treatment group sample sizes. Subjects were stratified according to their baseline categorical pain intensity (either moderate or greater than moderate) on a six-point rating scale described below.
For both patch and oral tablet treatment groups, active and nonactive study materials were identical in appearance. All patches and tablets were supplied in blank white pouches and high-density polyethylene bottles, respectively. A test kit containing the appropriate test materials was supplied to each subject. Each test kit was labeled with a two-part tear-off label containing study number, subject number, caution statements, product identification covered with a scratch-off surface (for breaking the mask in case of an adverse event), and other required information. The Clinical Supplies and Logistics department was responsible for maintaining the mask until after subject evaluability for efficacy analyses was determined and the database was locked.
Efficacy was measured by changes from baseline (reductions) in pain intensity and pain relief scores. Each woman completed a diary while enrolled in the study. Diaries contained a six-point categorical scale used to assess pain relief and a 101-point numerical rating scale (NRS-101) to indicate pain intensity.5 The NRS-101 scale, in which patients chose a number between 0 and 100, was anchored by directions that no pain = 0 and worst possible menstrual pain = 100. Initial pain intensity was recorded during a telephone contact (day 1) before treatment using the NRS-101 and a six-point categorical scale (none = 0, mild = 1, moderate = 2, moderately-severe = 3, severe = 4, extreme = 5) in which the patient had to indicate moderate or greater pain (2 or greater) to begin therapy. The scale used for pain relief has been reported previously.6 This scale is as follows: no relief = 0, a little relief = 1, less than half relief = 2, more than half relief = 3, a lot of relief = 4, and complete relief = 5. Pain intensity (NRS-101) and relief scores were recorded at hourly intervals to 6 hours after the start of therapy and again at treatment hours 8, 10, and 12 on day 1; hours 0, 2, 4, 6, 8, 10, and 12 on day 2; and upon waking the morning of day 3. A timer was used to prompt the recording of pain intensity and pain relief. Skin condition was assessed using a four-point categorical scale (normal color = 0, faint to definite pink = 1, definite redness = 2, intense redness = 3) before application of the abdominal patch, at the time of its removal on both treatment days, and upon waking the morning on day 3.
Quality of life was assessed before abdominal patch application on day 1 and before patch removal on day 2 with a rating instrument designed to assess premenstrual and menstrual symptoms and the severity of symptoms (absent, mild, moderate, and severe).7
Women whose menstrual flow did not begin within 24 hours of the onset of pain symptoms were declared not evaluable. Treatment failures were defined as the use of backup medication or treatment-related adverse events. In the event that a subject took backup medication or dropped out because of a treatment-related adverse event, her pain relief scores were set equal to zero for the remainder of the study, whereas her pain intensity scores were carried forward from the prior time point or set equal to baseline (whichever was more severe) for the remainder of the study.
The statistical methodology was done as follows. Reduction in pain intensity scores were calculated for each patient at each time point by taking differences from the patient’s baseline score. For pain relief and reduction in pain intensity, overall weighted means (adjusted for time) were then taken by subject for dose 1 data (0–6 hours), day 1 data (0–12 hours), and day 2 data (0–12 hours). The mean of the weighted means for days 1 and 2 was also calculated as an overall study measure. Overall weighted means were calculated as follows:
 |
where yi represents the value for pain relief or reduction in pain intensity at the ith posttreatment time point, and ti represents the value of time (in hours) at the ith posttreatment time point.
Reduction in pain intensity scores was analyzed by analysis of covariance procedures, adjusted for baseline. Pain relief scores were analyzed by one-way analysis of variance. For both pain intensity and pain relief, tests for treatment effects were conducted by time point and for the dose 1, day 1, day 2, and days 1 and 2 overall weighted means via independent-sample t tests on the least squares means. Overall 2-day mean results were corroborated by nonparametric one-way analysis of variance methods8–10 using Cochran-Mantel-Haenszel tests and Wilcoxon rank-sum tests for pain intensity reduction and pain relief, respectively. Repeated-measures, one-way analysis of variances were conducted on both variables to further characterize the efficacy profile over time. In addition, within the heated patch groups, Pearson and Spearman correlation coefficients were constructed to test for an association of pain relief and pain intensity reduction with the change in the abdominal skin temperature as measured at the third visit (afternoon, first day of treatment).
The onset of relief was determined using a stopwatch, as the time at which the subject experienced noticeable pain relief. Onset times were compared between treatment groups using Wilcoxon tests. In addition, the incidence rates of subjects reporting complete relief (pain relief score = 5) during the 2 days of treatment were compared between groups using Wald 
2 tests from a logistic regression model with terms for treatment group.
For quality-of-life data, symptom scales were summed into four symptom clusters (negative affect, water retention, food, and pain, as well as overall) by subject. Changes from baseline scores were determined by question, cluster, and overall for each subject. Analysis of covariance was done for these data by question, cluster, and overall, by examining effects for baseline and treatment. Pair-wise treatment differences were detected by independent-sample t tests on the least squares means. Results were corroborated by nonparametric one-way analysis of variance using Cochran-Mantel-Haenszel tests.
Subjective skin quality assessment (color change) scores were summarized by time point. The distribution of these data was compared between heated and unheated patch groups with Fisher exact tests.
Statistical significance was chosen at the P 
.05 level for this study. For pain intensity, pain relief, and quality-of-life data, t tests comparing heated patch plus placebo and unheated patch plus ibuprofen were two-tailed; all other pair-wise treatment comparisons were one-tailed, according to the protocol. Nonparametric efficacy results have been presented only in cases where different conclusions from its parametric counterpart were reached (eg, significant compared with nonsignificant).
Because no previous data were available to suggest the magnitude of change that could be expected, the choice of sample size was not based on statistical power for pain relief or other variables. A post-hoc power analysis was conducted on the 2-day pain relief data. We found that there was at least 85% power to detect a large, true difference of one unit in the 2-day pain relief treatment means for comparisons with the unheated patch plus oral placebo group using a one-tailed test at the .05 level of significance, based on an observed within-group standard deviation of 1.147.
|
Results |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The study groups each had 20 subjects, except for the unheated patch plus ibuprofen group, which had 21 subjects. There were no differences between groups in age, height, weight, abdominal girth, pretreatment skin temperature, or initial pain severity (Table 1
).
|
). Moreover, there was faster improvement in pain relief on initiation of treatment when heat was applied concomitantly with ibuprofen than when ibuprofen was used alone. The patient-recorded time to onset of noticeable pain relief was statistically significantly lower for patients in the heated patch plus ibuprofen group (median 1.5 hours) compared with the unheated patch plus ibuprofen group (median 2.79 hours, P = .01). Complete relief during the 2 days of study treatment relative to the unheated patch plus placebo group (35% incidence) was significantly higher for the heated patch plus ibuprofen group (68% incidence, odds ratio [OR] 4.0, P = .02) and the heated patch plus placebo group (70% incidence, OR 4.3, P = .015) but not for the unheated patch plus ibuprofen group (55% incidence, OR 2.3, P = .103). These results indicated that the odds of receiving complete relief from heat treatment were four times greater than they were for unheated patch plus placebo treatment.
|
).
Similar results were obtained for reduction of pain intensity (Figure 2). Overall for the 2 study days, significant differences were found for comparisons of the unheated patch plus placebo group compared with the heated patch plus placebo group, unheated patch plus ibuprofen group, and the heated patch plus ibuprofen group. Specifically, day 1–2 pain intensity reductions for the heated patch group (mean 40.4), the ibuprofen group (mean 39.0), and the heated patch plus ibuprofen group (mean 43.8) were significantly greater than that for the unheated patch plus placebo group (mean 21.9, P < .003). (For reference, the mean baseline pain intensity score was 69.2 of 100 points). These represented observed additional 84.5%, 78.1%, and 100% decreases in pain intensity, respectively, relative to the unheated patch plus placebo group. However, no overall day 1–2 significant difference was detected between the heated patch plus placebo group and unheated patch plus ibuprofen group themselves (P = .8). In addition, no difference in reduction of pain intensity was observed between the heated patch plus ibuprofen group and the unheated patch plus ibuprofen group during days 1–2 (P = .22).
|
.05 for all four statistics), although the main effect for treatment was still significant (P = .006). By analyzing successive time point differences, we found that the time profile was significantly altered twice on day 2: between hours 0 and 2 (P = .01) in a manner similar to that of pain relief and between hours 10 and 12 (P = .045). At hour 12, the heated patch plus ibuprofen treatment began to decline in efficacy, whereas all other treatment groups slightly increased reduction in pain intensity (Figure 2). Pearson correlation coefficients suggested only a weak correlation between pain relief and change in skin temperature for both the heated patch plus ibuprofen (r = .38, P = .108) and heated patch plus placebo (r = .32, P = .168) groups. Additional sensitivity was obtained by combining the data from these two treatment groups, producing a significant Pearson correlation coefficient (r = .36, P = .024 [Pearson]; r = 0.31, P = .058 [Spearman]). There were no significant correlations of pain intensity reduction and skin temperature in individual treatment groups, and only marginal significance (r = .31, P = .052 [Pearson]; r = 0.26, P = .104 [Spearman]) was achieved when the heat patch groups were combined for analysis.
For quality-of-life data, the heated patch plus placebo group had significantly greater reduction in overall symptom severity scores compared with the unheated patch plus placebo group (means = 0.42 compared with 0.22, respectively, P = .021). Overall symptom severity scores for the unheated patch plus ibuprofen group (mean = 0.36, P = .08) and the heated patch plus ibuprofen group (mean = 0.31, P = .18) were not significantly different from that of the unheated patch plus placebo group. For reference, the mean overall symptom severity score at baseline was 1.06. The heated patch plus placebo group had a significantly greater reduction in a specific quality-of-life symptom (tension) than did either the unheated patch plus ibuprofen group (P = .016) or unheated patch plus placebo group (P = .002). In addition, the heated patch plus placebo group had a significantly greater reduction in the specific quality-of-life symptom headache (P = .008) and the symptom cluster for water retention (P = .043 [t test]; P = .058 [Cochran-Mantel-Haenszel test]) compared with the unheated patch plus placebo group. Differences between the heated patch plus placebo and the unheated patch plus placebo groups in other specific symptoms or clusters on the quality-of-life assessment were not significant, but severity reductions in cramping, tenderness or fullness of breasts, and the cluster negative affect approached significance (.05 < P < .10).
Average skin temperature under the active heat patch increased from 32.8C before treatment to 37C after 6–8 hours of therapy (change of 4.2C). The mean change in skin temperature for the unheated patch groups was 1.4C. Among subjects who had a heated patch, 17 of 40 (42.5%) reported pinkness on day 1 after 12 hours of wear compared with five of 41 (12.2%) subjects who had an unheated patch. In addition, at this time point, two subjects from the heated patch group reported redness compared with one subject in the unheated patch group. The incidence rates of pinkness or redness were significantly different between subjects who had heated and unheated patches (Fisher exact test, P = .002). By the next morning (day 2, hour 0), all but one subject indicated that their skin had returned to normal; that subject, who was assigned to a heated patch group, reported residual pinkness (P = .49). After an additional 12 hours of patch wear on day 2, 23 of 40 (57.5%) subjects who had a heated patch reported pinkness compared with five of 41 (12.2%) subjects who did not have a heated patch. Only one subject in a heated patch group reported redness at that time point (P < .001). By the next morning (day 3, hour 0), all but two subjects indicated that their skin had returned to normal; both subjects, who were assigned to a heated patch group, reported residual pinkness (P = .24). All patients reported normal skin at the exit visit (day 3–7). There were no drug or heat-related adverse effects during this study.
|
Discussion |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Our results also raise some interesting questions about the mechanism by which relief is attained. Inspection of the data in Figures 1 and 2 suggests that, although overall there is an additional effect when both heat and ibuprofen were used together, the incremental effect was less than that achieved by either modality separately. This suggests that both treatment modalities might act through similar mechanisms, eg, by a central analgesia effect, through a direct effect on the activity of the uterus, or both. Nonsteroidal anti-inflammatory agents such as ibuprofen are known to have activity as central analgesics (in addition to their direct action on uterine contractile activity), but the mechanism of this action is largely unknown. Heat could act as an analgesic through a form of gate control (similar to trans-cutaneous electrical nerve stimulation) or by altering pain thresholds centrally or through an altered sense of well being.11 This mechanism might mimic the changes seen during placebo therapy.12 Similarly, nonsteroidal anti-inflammatory agents have been shown to alter uterine activity in primary dysmenorrhea.13–15 Heat applied to the anterior abdominal wall has been shown to decrease the activity of the gastrointestinal tract,16 suggesting the possibility of a feed-back mechanism that could have direct uterine relaxing effects. Warming the skin of the lower thorax also has been reported to increase intestinal blood flow in rats.17 The relative contribution from these potential actions of topical heat on the uterus awaits clarification.
The topical heat treatment used in this study appeared to have an additive effect with ibuprofen, as the time to onset of pain relief was significantly shorter with the combination of heated patch plus ibuprofen therapy compared with ibuprofen treatment plus unheated patch. In addition, the pain relief in the heated patch plus ibuprofen group on the first day of treatment was directionally higher than that in the unheated patch plus ibuprofen group.
Subjects who had a heated patch plus placebo showed continuing pain relief at time zero on day 2 of therapy, and this pain relief was statistically different from the unheated patch plus placebo group (P = .002). This statistically significant next-day pain relief (day 2) was not present for subjects who used ibuprofen, either singly or in combination with the application of heat. The absence of a reciprocal increase in pain relief score for women who had unheated patch plus placebo treatments suggests stability in the degree of dysmenorrhea that does not change until the start of day 3, when spontaneous resolution might be expected to begin. This extended pain relief on the next day after heat treatment supports the hypothesis of a centrally altered pain threshold, although this mechanism remains conjectural.
|
Footnotes |
|---|
Financial Disclosure
Dr. Akin is a contractual employee of Health Quest Therapy and Research Institute, and authors Weingand, Hengehold, Goodale, and Hinkle are employees of Procter and Gamble.
Received April 19, 2000. Received in revised form September 8, 2000. Accepted September 13, 2000.
|
References |
|---|
|
Top
AbstractMaterials and MethodsResultsDiscussionReferences |
|---|
2. Lehman JF, de Lateur BJ. Ultrasound, shortwave, microwave, laser superficial heat, and cold in the treatment of pain. In: Wall PD, Melzack, eds. Textbook of pain. 3rd ed. Edinburgh: Churchill Livingstone, 1994:1239.
3. Vance AR, Hayes SH, Spielholz NI. Microwave diathermy treatment for primary dysmenorrhea. Phys Ther 1996;76:1003–8.
4. Spilker B. Guide to clinical trials. New York: Raven Press, 1991; 15–20.
5. Jensen MP, Karoly P, Braver S. The measurement of clinical pain intensity: A comparison of six methods. Pain 1986;27:117–26.[Medline]
6. Weingand K, Hengehold D, Knight E, Hinkle R, Combs L, McNutt B, et al. Topical heat provides pain relief of delayed onset muscle soreness of the distal quadriceps muscles. Med Sci Sports and Exerc 1999;31:S75.
7. Thys-Jacobs S, Alvir JMJ, Fratarcangelo P. Comparative analysis of three PMS assessment instruments—The identification of premenstrual syndrome with core symptoms. Psychopharmacol Bull 1995; 31:389–96.[Medline]
8. Quade D. Rank analysis of covariance. J Am Stat Assoc 1967;62: 1187–200.
9. Koch GG, Amara IA, Davis GW, Gillings DB. A review of some statistical methods of covariance analysis of categorical data. Biometrics 1982;38:563–95.[Medline]
10. Koch GG, Carr GJ, Amara IA, Stokes ME, Uryniak TJ. Categorical data analysis. In: Berry DA, ed. Statistical methodology in the pharmaceutical sciences. New York: Wiley, 1985:128–33.
11. On AY, Colakoglu Z, Hepguler S, Aksit R. Local heat effect on sympathetic skin responses after pain of electrical stimulus. Arch Phys Med Rehabil 1997;78:1196–9.[Medline]
12. Smith RP. Objective changes in intrauterine pressure during placebo therapy of primary dysmenorrhea. Pain 1987;29:59–66.[Medline]
13. Smith RP, Powell JR. Intrauterine pressure changes during mefenamic acid treatment of primary spasmodic dysmenorrhea. Am J Obstet Gynecol 1982;143:286–92.[Medline]
14. Smith RP, Powell JR. Simultaneous objective and subjective evaluation of meclofenamate sodium in the treatment of primary dysmenorrhea. Am J Obstet Gynecol 1987;157:611–6.[Medline]
15. Smith RP. The dynamics of nonsteroidal anti-inflammatory therapy for primary dysmenorrhea. Obstet Gynecol 1987;70:785–8.[Abstract]
16. Fischer E, Solomon S. Physiologic responses to heat and cold. In: Licht S, Kamenetz HL, eds. Therapeutic heat and cold. 2nd ed. New Haven, Connecticut: Elizabeth Licht, 1965:126–69.
17. Kuntz A. Anatomic and physiologic properties of cutaneous-visceral vasomotor reflex arcs. J Neurophysiol 1945;8:421–9.
This article has been cited by other articles:
|
|
 |
|
  M. Proctor and C. Farquhar Diagnosis and management of dysmenorrhoea. BMJ, May 13, 2006; 332(7550): 1134 - 1138. [Full Text] [PDF]
|
|
|
|
 |
|
 L. Nasir and E. T. Bope Management of Pelvic Pain from Dysmenorrhea or Endometriosis J Am Board Fam Med, November 1, 2004; 17(suppl_1): S43 - S47. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. R Kessenich Continuous topical heat was as effective as ibuprofen for dysmenorrhoea Evid. Based Nurs., October 1, 2001; 4(4): 113 - 113. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
).