Clinical Safety Assessment of an Ultra Absorbency Menstrual Tampon

Abstract

Background:

Menstrual tampons are available in a range of absorbencies to allow women to use the product most appropriate to their needs. This study assessed the safety of an ultra absorbency (15 g to 18 g) tampon compared with a currently marketed super-plus absorbency (12 g to 15 g) tampon as a control.

Methods:

Healthy women age 18–45 years (n = 95) were enrolled in this single-blind, crossover study. Subjects used, in random order, the experimental tampon during one menstrual cycle and the control tampon during the other. Subjects were also randomly assigned to receive either vaginal microbial assessments for determination of the presence and density of Staphylococcus aureus (n = 35) or colposcopic examinations for assessment of changes in the vaginal mucosa (n = 60). Data on comfort and acceptability of the tampons were collected by using diaries and questionnaires completed by the subjects in both groups.

Results:

Twenty-eight women in the microbial assessment group and 43 in the colposcopic examination group completed the study; these subjects made up the primary analysis population. No differences in isolation frequency or cell density of S. aureus or in vaginal mucosal changes were observed with the experimental tampon in comparison with the control tampon. No reported adverse events were attributed to the experimental tampon. Both tampons received positive comfort ratings.

Conclusions:

Based upon microbial assessments, colposcopic examinations, adverse events, and subject reporting of comfort, the ultra absorbency tampon is similar in safety profile to the currently marketed super plus absorbency tampon.

Introduction

The use of vaginal tampons to absorb menstrual flow has a long history in the United States, where a tampon product was patented and became popular more than seventy years ago; safety assessments of tampons date back to the 1940s.^1
–3

As a medical device, tampons are regulated by the Food and Drug Administration (FDA), which has established standardized terms to be used to identify the absorbency of the product.⁴ The designated terms used to describe tampon absorbency range from “light,” for tampons that absorb 6 g or less of saline in testing, to “ultra,” for tampons that absorb 15 g to 18 g of saline. Because menstrual flow varies from woman to woman and over the course of a menstrual cycle, each woman may need a variety of tampons with different absorbencies to manage her menstrual flow, and a product with ultra absorbency is needed by some women when they are having heavy flow. Standardized terms to describe absorbency are included in the product labeling to enable the consumer to make comparisons among products and to allow her to select the minimum absorbency tampon needed to control menstrual flow, in order to minimize the risk of toxic shock syndrome.⁴

Following the identification of an association between tampon use and toxic shock syndrome, many researchers have evaluated factors that may play a role in risk, including effects of tampon composition on growth of S. aureus or production of toxic shock syndrome toxin-1,^5,6 tampon-induced vaginal injury or irritation,^7
–9 and tampon absorbency.^10,11 All tampons currently marketed in the United States are made of cotton, rayon, or cotton/rayon mixtures, which do not amplify in vitro production of toxic shock syndrome toxin-1 by S. aureus. ³

Ultra absorbency tampons have only recently been made available to women, and there are virtually no data describing the safety of their use. This study compares the safety profile of a new ultra absorbency tampon product with a currently marketed super plus absorbency tampon by assessing the effects of each product on the vaginal microbiota (isolation frequency and cell density of S. aureus) and vaginal mucosa. Super plus absorbency tampons were selected as the control product because they were the highest absorbency tampon available prior to the introduction of ultra absorbency tampons. The data from this study will help to define the safety profile of this new tampon option for menstruating women.

The terminology we use in this manuscript (i.e., ultra, super plus) is per the absorbency classification for tampons (nomenclature and labeling) as defined in the Code of Federal Regulations (21 CFR 801.430) by the FDA. Tampons are a class 2 medical device regulated by the FDA. Labeling of absorbency using these terms is a requirement of all tampon manufacturers to assure consistency across brands.

Materials and Methods

Study design

This study was a single-center, randomized, controlled, crossover study conducted over 2 menstrual cycles to evaluate the safety of an ultra absorbency (15 g to 18 g) tampon under development compared with a currently marketed super plus absorbency (12 g to 15 g) tampon (Tampax Pearl^® Super Plus) as a control. Each subject used either the experimental tampon or the control tampon during the first cycle and the other study product during the second cycle. Although the women were aware of the absorbency of the study tampons, microbiological and colposcopic assessments were performed by individuals who were blinded to the treatment sequence.

Subjects

Healthy women aged 18 to 45 years were eligible to enroll if they had regular menstrual cycles, used tampons exclusively and normally used 3 or more super plus (12 g to 15 g) or ultra (15 g to 18 g) absorbency tampons during their periods, and had had a gynecological exam (including pap smear) within the past two years. They also agreed to use only menstrual products (study tampons and pads to be used before and after study tampon use during each cycle until their study visit) provided by the Sponsor, and to refrain from the use of anti-inflammatories, antihistamines, steroid medications, and vaginal products (including douches, vaginal medications, feminine deodorant sprays, etc.) during the study, and refrain from sexual intercourse for 24 hours prior to visits at which microbial evaluations or colposcopic examinations were to be conducted.

Women were excluded from participation if they had menstrual irregularity, were pregnant or intended to become pregnant, had a history of toxic shock syndrome or heart valve replacement, had a positive pap smear within the last 2 years, had taken anti-inflammatories, antihistamines, or steroid medications within the last 7 days, had a medical condition that might compromise the immune system or had a history of immunosuppressive drug therapy, chemotherapy, or radiation therapy, had taken antibiotics or antifungal medications in the last 4 weeks, had diabetes, a vulvar piercing, or history of genital herpes. Women with vaginal irritation, abrasions, or lesions, or who had active vaginal infection at the screening visit were also excluded. All women gave written informed consent prior to entry into the study. The study protocol was approved by an Institutional Review Board prior to initiation of the study.

Study procedures

At the screening visit, women gave written informed consent, were screened for inclusion and exclusion criteria, provided gynecologic history and information about medications, and underwent microbial assessment (for bacterial vaginosis [BV], trichomoniasis, yeast infection, gonorrhea, and genital chlamydia), colposcopic examination, and vaginal assessment. Following verification of eligibility, subjects returned to the study site for a visit at which study products, use diaries, and comfort questionnaires were distributed, and subject instructions were reviewed. Subjects were instructed to use the study tampons in place of their usual super plus or ultra absorbency tampons during the next 2 menstrual cycles. Subjects were randomly assigned to receive either microbiologic assessments for the isolation frequency and cell density of S. aureus (study procedures Group 1) or colposcopic examinations (study procedures Group 2) during the study. A total sample size of 95 women (35 in Group 1 and 60 in Group 2) was estimated to be required to ensure at least 75 women (25 in Group 1 and 50 in Group 2) would complete the study. Women were randomly assigned to treatment sequence. Women were asked to return to the study center for examination between 3 and 48 hours after removal of the last study tampon during each cycle.

Study tampons

The experimental and control tampons were distributed in plain white wrappers and packaged in clear plastic bags labeled with the product absorbency and the required toxic-shock-syndrome information for tampon products. Both tampons contained 100% rayon in the absorbent pad; the only difference between the 2 products was the absorbency.

Screening microbiological procedures

Microbial assessments were completed on samples from all subjects at screening for identification of bacterial vaginosis, trichomoniasis, yeast infection, gonorrhea, and genital chlamydia. Subjects were excluded from the study if any of these diseases were identified.

A sterile dacron swab (Copan Diagnostics, Corona, CA) was used to obtain vaginal specimens for wet mount analyses. Cervical samples for gonorrhea and chlamydial infection were obtained using a Gen-Probe (Gen-Probe, San Diego, CA) specimen collection kit for endocervical specimens.

Bacterial vaginosis was determined by the presence of three of the four Amsel criteria diagnostic of BV.¹² If the presence of menses precluded the use of pH and discharge as diagnostic criteria, the diagnosis of BV was made based on the presence of fishy amine odor upon addition of potassium hydroxide to the vaginal sample and the presence of clue cells. Trichomoniasis was confirmed based on the presence of motile trichomonads in a wet mount preparation within 15 minutes of sampling.¹³ Yeast infection was diagnosed if pseudohyphae forms could be identified.¹⁴ Diagnosis of chlamydia and gonorrhea was determined using a rapid-DNA probe for Chlamydia trachomatis and Neisseria gonorrhoeae (Pace 2, Gen-Probe, San Diego, CA).

S. aureus microbiological assessment

The frequency of isolation of S. aureus and cell density were determined from vaginal swab samples obtained from the mid-vagina region of subjects enrolled in Group 1. Care was taken to avoid contact with surfaces other than those designated for sampling. Swabs were weighed (to 0.xxx g) before and after sample collection. Quantitative results were reported as colony forming unit (CFU) per gram of vaginal sample. Representative colonies were confirmed as S. aureus via subculturing on 5% sheep blood agar and for subsequent gram stain, catalase test, and tube coagulase and BBL crystal gram positive ID tests (Becton Dickinson, Franklin Lakes, NJ).

Colposcopic examination

Colposcopic examination was performed on all subjects at screening and on subjects in Group 2 during the 2 menstrual cycles. This examination included assessment of the labia minora, introitus, lower and middle vaginal walls, upper vagina including fornices, and cervix for potential erythema (on a 0–4 grading scale), abrasions, or ulceration.

Other assessments

All adverse reactions reported during the study were recorded. Subjects also completed a tampon use diary where they recorded times of insertion and removal of the study tampon, a rating of overall comfort, and notes of any discomfort associated with the tampon use. Subjects also rated the tampon overall (ranging from excellent to poor) and its comfort (ranging from very comfortable to very uncomfortable), and provided voluntary comments on a monthly comfort questionnaire.

Statistical methods

Summary statistics for age were calculated by group and treatment sequence to assess overall balance between groups and between treatment sequences. Generalized estimating equations (GEEs)¹⁵ using a logit link and exchangeable correlation structure were used to analyze binary response variables. GEEs using a cumulative logit link and independent correlation structure were used to analyze ordinal response variables. Two-period crossover analysis methods proposed by Grizzle¹⁶ and Koch¹⁷ were used to analyze continuous variables. Cochran-Mantel-Haenszel tests also were conducted. These include McNemar's test,¹⁸ conducted to compare treatment effects on erythema incidence greater than 0, erythema incidence greater than 0.5, and isolation of S. aureus.

Results

Analysis population

Ninety-five subjects met the study criteria and were enrolled into the study; 35 in Group 1 (microbiological assessment) and 60 in Group 2 (colposcopic assessments). A total of 13 subjects (6 in Group 1; 7 in Group 2) did not complete the study, 1 subject (Group 2) was improperly enrolled in the study, and 10 additional subjects (1 in Group 1 and 9 in Group 2) took prohibited medications during the treatment phase. These patients were excluded from the primary analysis population. Questionnaire and diary data from 3 subjects (2 in Group 1 and 1 in Group 2) who did not complete the study were included in the primary statistical analyses because their questionnaire or diary data were considered valid, despite the subject's not having completed the final clinic visit within the prespecified time window. Therefore, of the 35 subjects who enrolled in Group 1, 28 subjects made up the primary analysis population, while data from 29 subjects were included in the analysis of the questionnaire data, and data from 30 subjects were included in the analysis of the diary data. Of the 60 subjects enrolled in Group 2, 43 subjects made up the primary analysis population, while data from 43 subjects were included in the analysis of questionnaire data and data from 44 subjects were included in the analysis of diary data. Results of secondary analyses including data from women who took prohibited medications were consistent with those of the primary analyses.

The mean age of the subjects was 34 years (minimum, 19; maximum, 45), with no apparent differences between groups based on treatment sequence or assignment to Group 1 or Group 2.

Extent of exposure

All subjects used at least two each of both the experimental and control tampons. There were no statistically significant differences in usage of the products based upon average tampon wear time, average daily number used, total number of tampons used, total cumulative tampon wear time, and percent of tampons worn longer than 8 hours (Table 1).

Table 1.

Tampon Usage

	Experimental tampon		Control tampon
Parameter	Mean	SE	Mean	SE	p-value
Wear time (hrs)
Group 1	4.7	0.24	4.8	0.31	0.616^a
Group 2	5.5	0.25	5.4	0.21	0.757^b
Number used per day
Group 1	3.1	0.11	3.2	0.20	0.704^a
Group 2	2.8	0.13	2.7	0.12	0.949^b
Daily cumulative wear time (hr)
Group 1	14.2	0.68	14.3	0.69	0.582^a
Group 2	15.2	0.65	13.8	0.52	0.014^a
Total number used
Group 1	8.2	0.36	8.2	0.39	0.612^b
Group 2	8.3	0.31	8.2	0.32	0.628^b
Total cumulative wear time (hr)
Group 1	38.6	2.62	37.6	2.42	0.320^a
Group 2	45.1	2.18	43.7	2.35	0.237^a
% tampons with wear time > 8 hr
Group 1	5.9	1.73	9.7	2.90	0.926^b
Group 2	9.5	2.49	12.3	2.94	0.812^b

Group 1, Microbial group, n = 30; Group 2, Colposcopy group, n = 44.

Statistical tests (paired t-test or signed rank test) were determined using methods proposed by Grizzle¹⁶ and Koch.¹⁷

One-sided p-value based on paired t-test.

One-sided p value based on signed rank test.

SE, standard error.

The one statistically significant difference noted was the average daily cumulative wear time in the Group 2 subjects, who used the study tampons for more hours of the day than they did the control tampons. No similar difference was seen in Group 1 subjects.

S. aureus microbiological assessments

S. aureus was isolated from 7 of the 28 subjects in Group 1. S. aureus was isolated in 3 subjects only after the experimental tampon cycle, in 3 subjects only after the control tampon cycle, and in 1 subject after both cycles. There was no statistically significant difference in isolation frequency between the treatment cycles (McNemar's χ-square test, p = 0.327). When present, the S. aureus cell density ranged from 9 × 10³ CFU/gram to >1.9 × 10⁷ CFU/gram of vaginal fluid sample.

Colposcopic assessments

No abrasions or ulcerations were observed during the course of the study. Erythema scores when observed were usually 0.5 (faint, barely perceptible), with an occasional score of 1.0 (faint but definite) on a scale of 0–4 (Table 2). There was no statistically significant difference in distribution of the erythema scores between the experimental and control tampons.

Table 2.

Number and Percent of Patients with Erythema Scores ≥0

	Experimental tampon			Control tampon
	Erythema Score			Erythema Score
Location	0	0.5	1.0	0	0.5	1.0	p value
Labia minora	18 (42%)	19 (44%)	6 (14%)	11 (26%)	21 (49%)	11 (26%)	0.958
Introitus	32 (74%)	11 (26%)	0	23 (54%)	18 (42%)	2 (5%)	0.983
Lower vaginal wall	36 (84%)	7 (16%)	0	26 (61%)	17 (40%)	0	0.996
Middle vaginal wall	34 (79%)	9 (21%)	0	28 (65%)	15 (35%)	0	0.946
Upper vagina	35 (81%)	8 (19%)	0	26 (61%)	17 (40%)	0	0.994
Cervix	30 (70%)	11 (26%)	2 (5%)	24 (56%)	17 (40%)	2 (5%)	0.910

One-sided p value is based on the Cochran-Mantel-Haenszel (CMH) test for row mean scores using rank scores (distribution of of scores). Generalized estimating equation analysis was conducted on erythema incidence only, with results that support the CMH p values.

Adverse events

Ten subjects reported adverse events during the study. One subject had mild superficial labial irritation during use of the control tampon, which resolved without treatment and was considered to be possibly related to tampon use. The remaining 9 reported adverse events (head cold, 4; nasal congestion, 1; bacterial vaginosis, 1; seasonal allergies, 1; migraine headache, 1; scratched left cornea, 1) were judged unrelated to the use of either the experimental or control tampons.

Product acceptance assessments

Tampon experience data from the diary indicated satisfaction with the study products, based on the favorable comfort scores, the low percentage of tampons associated with insertion, wear, or removal discomfort, and the low incidence of burning, stinging, itching sensations (Table 3). None of the differences in subject ratings between the experimental and control tampons were statistically significant.

Table 3.

Tampon Use Diary Summary

	Experimental tampon			Control tampon
	Mean	SE	Median	Mean	SE	Median	p value
Comfort Scores
Overall
Group 1	1.0	0.17	1.0	1.3	0.13	1.7	0.212^a
Group 2	0.9	0.11	1.0	1.0	0.11	1.0	0.219^a
Minimum overall comfort ^c
Group 1	0.5	0.24	1.0	0.7	0.20	1.0	0.440^b
Group 2	0.2	0.16	0.0	0.3	0.17	1.0	0.435^a
% Tampons associated with
Any symptom
Group 1	20.6	5.5	0.0	19.4	5.44	0.0	0.721
Group 2	14.0	2.9	10.0	16.6	3.3	10.0	0.718
Any sensation
Group 1	5.3	2.8	0.0	1.7	1.7	0.0	0.937
Group 2	0.7	0.6	0.0	0.6	0.4	0.0	0.500
Insertion discomfort
Group 1	3.0	1.7	0.0	2.0	1.0	0.0	0.758
Group 2	4.0	1.2	0.0	5.6	1.6	0.0	0.688
Wearing discomfort
Group 1	7.0	2.6	0.0	2.6	1.4	0.0	0.926
Group 2	5.8	2.1	0.0	6.1	2.4	0.0	0.526
Removal discomfort
Group 1	12.1	4.7	0.0	15.2	5.2	0.0	0.712
Group 2	6.7	1.9	0.0	9.8	2.7	0.0	0.934
Burning sensation
Group 1	4.1	2.3	0.0	0.6	0.6	0.0	0.875
Group 2	0.0	0.0	0.0	0.2	0.2	0.0	0.500
Stinging sensation
Group 1	1.1	1.1	0.0	0.0	0.0	0.0	0.500
Group 2	0.2	0.2	0.0	0.3	0.3	0.0	0.500
Itching sensation
Group 1	1.8	1.3	0.0	1.1	1.1	0.0	0.500
Group 2	0.5	0.5	0.0	0.0	0.0	0.0	0.500

Group 1, Microbial group, n = 30; Group 2, Colposcopy group, n = 44.

Comfort score rating: −2, −1, 0, 1, 2 (very uncomfortable, uncomfortable, neither comfortable nor uncomfortable, comfortable, and very comfortable).

Statistical tests (paired t-test or signed rank test) were determined using methods proposed by Grizzle¹⁶ and Koch.¹⁷ The generalized estimating equation (GEE) did not always converge, so Grizzle and Koch two-period crossover results are reported. The GEE results support the reported results.

One-sided p value is based on signed rank test unless otherwise noted.

Two-sided p value is based on signed rank test.

Two-sided p value is based on paired-t test.

Mean lowest score reported.

SE, standard error.

The overall rating and comfort ratings from the monthly comfort questionnaires were very high for both the experimental and control tampons. The mean overall combined group rating score was 3.9 for the experimental tampon and 4.0 for the control tampon (on a scale of 1–5, with 1 being poor and 5 being excellent). These results were consistently very positive for both experimental and control tampons, with no statistically significant differences (p ≥ 0.209) in the distribution of rating scores for either the overall rating or the comfort rating (Table 4).

Table 4.

Monthly Comfort Questionnaire Summary

	Experimental tampon						Control tampon
Overall Rating
	Poor	Fair	Good	Very good	Excellent	Poor	Fair	Good	Very good	Excellent	p value
Group 1	1 (3%)	2 (7%)	8 (28%)	7 (24%)	11 (38%)	0 (0%)	3 (10%)	4 (14%)	10 (34%)	12 (41%)	0.376
Group 2	0 (0%)	2 (5%)	11 (26%)	17 (40%)	13 (30%)	1 (2%)	3 (7%)	7 (16%)	19 (43%)	13 (30%)	0.876
Comfort Rating
	−2	−1	0	1	2	−2	−1	0	1	2
Group 1	0 (0%)	3 (10%)	6 (21%)	6 (20%)	14 (48%)	0 (0%)	2 (7%)	1 (3%)	12 (41%)	14 (48%)	0.209
Group 2	1 (2%)	0 (0%)	8 (19%)	20 (47%)	14 (33%)	0 (0%)	2 (5%)	7 (16%)	16 (37%)	18 (42%)	0.572

Group 1, Microbial group, n = 29; Group 2, Colposcopy group, n = 43. Values are the numbers and percentage of subjects giving each score.

Comfort score rating: −2, −1, 0, 1, 2 (very uncomfortable, uncomfortable, neither comfortable nor uncomfortable, comfortable, and very comfortable).

p values are based on the stratified Cochran-Mantel-Haenszel test for row mean scores (2-sided test). Generalized estimating equation (GEE) results support the reported results.

Discussion

This study found no differences between the ultra absorbency tampon and the currently marketed super plus absorbency tampon with regard to effects on vaginal microflora (presence and density of S. aureus), changes in vaginal mucosa, or subject ratings of comfort and acceptability. These results suggest that the products will have similar favorable safety profiles when used by consumers.

Tampons sold in the United States are accompanied by consumer information materials that alert the consumer to the association between tampon use and menstrual toxic shock syndrome, describe the symptoms of the condition, and provide instructions to reduce risk, including not using tampons, using tampons alternately with sanitary napkins, and using the lowest absorbency tampon needed to control menstrual flow. Over the period from 1979 to 1996, dramatic decreases in the number of reported cases of menstrual toxic shock syndrome were observed in the United States.¹⁹ The comparison product in this study the Tampax Pearl^© Super Plus tampon, has been sold in the United States since 2002 and has an established, acceptable safety profile. In comparing the two products, the microbial assessment was important to determine whether the increase in absorbency of the test product would have any effect on the presence of or promotion of S. aureus growth, because some strains of this organism produce the toxic shock syndrome toxin. The absence of such effects is reassuring. Similarly, the experimental tampon was not associated with any increase in vaginal injury as visualized by colposcopy; any increase in irritation or erosion could have an effect on vaginal health. It is interesting that tampon use patterns, including average tampon wear time, average daily number used, total number of tampons used, total cumulative tampon wear time, and percent of tampons worn longer than 8 hours, did not differ between the two tampons. One might have expected that the higher absorbency could cause changes in use, such as longer wear times, but this was not seen. The similarity in use pattern also provides reassurance that the two tampons will be similar in their safety profiles. No clinically important adverse events were observed with either tampon during the limited period of the study. When asked to evaluate comfort of the tampons, subjects rated both products favorably, with no meaningful differences between the experimental and control tampons, suggesting the experimental tampon will be acceptable to consumers.

Conclusions

In this controlled study, the ultra absorbency tampon was similar in all safety assessments to the currently marketed super plus absorbency tampon.

Footnotes

Acknowledgments

The authors thank Stephen Nunn, PA-C for performing the subject examinations, Catherine Davis, Ron Berg, and Bette Wojack for microbiological support, and Lisa C. Bosch for assistance in the preparation of the manuscript.

Author's Disclosure Statement

This study was supported financially by the Procter & Gamble Company. Dr. Anne Hochwalt is an employee of Procter & Gamble. Ms. Jones is retired from Procter & Gamble. Ms. Meyer was previously employed by Procter & Gamble and is now employed by HillTop Research.

References

Thornton

. The use of vaginal tampons for the absorption of menstrual discharges. Am J Obstet Gynecol, 1943; 46:259–265.

Magid

, Geiger

. The intravaginal tampon in menstrual hygiene: a clinical study. Medical Record, 1942; 155:316–318.

Dickinson

. Tampons as menstrual guards. JAMA, 1945; 128:490–494.

Code of Federal Regulations 21CFR801.430 (homepage on the Internet). Silver Spring, Md: Food and Drug Administration(updated April 1 2008; cited December 4 2008)http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfcfr/CFRSearch.cfm?FR = 801.430

Parsonnet

, Modern

, Giacobbe

. Effect of tampon composition on production of toxic shock syndrome toxin-1 by Staphylococcus aureus in vitro. J Infect Dis, 1996; 173:89–103.

Schievert

. Comparison of cotton and cotton/rayon tampons for effect on production of toxic shock syndrome toxin. J Infect Dis, 1996; 172:1112–1114.

Berkeley

, Micha

, Freedman

, Hirsch

. The potential of digitally inserted tampons to induce vaginal lesions. Am J Obstet Gynecol, 1985; 66:31–35.

Raudrant

, Frappart

, De Haas

, Thoulon

, Charvet

. Study of the vaginal mucous membrane following tampon utilisation; aspect on colposcopy, scanning electron microscopy and transmission electron microscopy. Eur J Obstet Gynecol Reprod Biol, 1989; 31:53–65.

Raudrant

, Landrivon

, Frappart

, De Haas

, Champion

, Ecochard

. Comparison of the effects of different menstrual tampons on the vaginal epithelium: a randomised clinical trial. Eur J Obstet Gynecol Reprod Biol, 1995; 58:41–46.

10.

Berkley

, Hightower

, Broome

, Reingold

. The relationship of tampon characteristics to menstrual toxic shock syndrome. JAMA, 1987; 258:917–923.

11.

Kehrberg

, Latham

, Haslam

et al. Risk factors for staphylococcal toxic-shock syndrome. Am J Epidemiol, 1981; 114:873–878.

12.

Amsel

, Totten

, Spiegel

, Chen

, Eschenbach

, Holmes

. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. Am J Med, 1983; 74:14–22.

13.

Wang

. Trichomoniasis. Prim Care Update Ob Gyns, 2000; 7:148–153.

14.

Shehin

, Jones

, Hochwalt

, Sarbaugh

, Nunn

. Clinical safety-in-use study of a new tampon design. Infect Dis Obstet Gynecol, 2003; 11:89–99.

15.

Diggle

, Liang

, Zeger

. Analysis of longitudinal data, 1st. New York: Oxford University Press, 1994.

16.

Grizzle

. The two-period change-over design and its use in clinical trials. Biometrics, 1965; 21:467–480.

17.

Koch

. The use of non-parametric methods in the statistical analysis of the two-period change-over design. Biometrics, 1972; 28:577–581.

18.

Fleiss

. Statistical methods for rates and proportions. New York: John Wiley & Sons, 1981; 112–119.

19.

Hajjeh

, Reingold

, Weil

, Shutt

, Schuchat

, Perkins

. Toxic shock syndrome in the United States: surveillance update, 1979–1996. Emerg Infect Dis, 1999; 5:807–810.