Embed Size (px)
Citation preview
Clinical trial
Can an instruction video or palpation aid improve the effectiveness of breast
self-examination in detecting tumors? An experimental study
Richard von Georgi1,2, Franziska Thele1, Andreas Hackethal3, and Karsten Munstedt31Institute of Medical Psychology and Sociology of the Justus-Liebig-University, Giessen, Germany; 2Institute ofMusicscience, Justus-Liebig-University, Giessen, Germany; 3University Department of Obstetrics and Gynecology,Justus-Liebig-University, Giessen, Germany
Key words: breast self-examination, effectiveness, instruction video, palpation aid
Abstract
Introduction. Although large randomized trials have not shown benefits for breast self-examination (BSE), manyorganizations still endorse the practice. This study aimed to determine whether an instruction video or palpation aidimproved the effectiveness of BSE.
Methods. A total of 100 volunteers (50 men and 50 women) with no previous experience of BSE and/or clinicalbreast palpation were randomly assigned to the following intervention: instruction video seen versus not seen anduse of palpation aids versus aids not used. Participants completed psychological trait and previous knowledgequestionnaires before the intervention and/or breast examination. Examination was carried out on 24 different sizedsilicone breast forms, into 20 of which had been inserted tumor equivalents measuring 0.8–3.0 cm in diameter. Thedependent variable was the rate of accurate tumor detection – the mean sum of correct positive hits (CPH) – definedas tumors detected within 20 s. Mean CPH values were then analyzed in relation to the interventions and othervariables including gender and psychological measures (ANOVA and COVAR).
Results. Neither the instruction video nor interaction effects between gender and the instruction video influencedthe mean CPH value. Furthermore, the palpation aid was strongly and significantly associated with a reduceddetection (p=0.00003).
Conclusion. Using an instruction video or palpation aid did not improve the effectiveness of breast examinationin detecting tumors – a finding which supports results from large randomized studies. It is difficult to understandwhy BSE is still promoted by various groups.
Introduction
Breast cancer is the most common malignancy in womenworldwide and a leading cause of death from cancer,especially in younger women [1]. The average loss ofpotential lifetime associated with breast cancer is 21%[1]. In addition, breast cancer affects women’s quality oflife and is associated with appreciable direct and indirecteconomic costs [1]. Early detection not only improvesthe patients’ chances of survival, at the very least itreduces the direct costs related to breast cancer treat-ment. Thus, great importance has been placed on sec-ondary prevention through mammography, breastultrasonography, clinical breast examination and breastself-examination (BSE) [1]. However, there is consider-able controversy about the value of all major secondaryprevention measures.
Breast self-examination is an inexpensive and non-invasive procedure involving regular examination of thebreast. The fact that about 65% of women between 40
and 45 years of age detected their own lumps - andearlier investigators reported even higher rates - has ledto the assumption that regular BSE is extremely usefulin disease prevention, and it is promoted as such byseveral oncological societies [2]. However, analysis oflarger breast cancer case series has shown that tumorself-detection is attributed equally to routine BSE andaccidental findings of breast lumps [3].
Arguments against the usefulness of BSE in second-ary cancer prevention have come from two, large ran-domized trials and several observational studies [4–8].Both the randomized trials and two meta-analyses of thedata concluded that BSE has no benefit, and indeed maydo harm [4–8]. One observational study found no dif-ferences in cancer mortality between women who prac-ticed BSE and those who did not [9]. Finally, it wasshown that neither the practice of BSE nor a high fre-quency of BSE was associated with a decreased risk oflate-stage disease; only those women who were veryproficient at BSE seemed to have a reduced mortality
Breast Cancer Research and Treatment (2006) 97: 167–172 � Springer 2005DOI 10.1007/s10549-005-9106-7
[10,11]. However, the retrospective design and theassociated recall bias in many related studies mean thatwe should exercise caution in interpreting the data[10,11].
Supporters of BSE have criticized the trials on self-examination. They argue that only 56% of the women inthe Russian study of Semiglazov et al. [8] performedBSE more than five times per year, and this is regardedas insufficient [1,11]. Furthermore, the women in theChinese studies of Thomas et al. [6,7] may not have beenproficient enough at the technique to detect anyabnormalities in their breasts. Supporters believe thatthe results would have been more positive for well-instructed and highly motivated women [1,11].
One probable result of this controversy is that BSEby itself is not generally promoted, except for womenwho have attended special instruction courses, seen vi-deo instruction tapes, or have been helped by the use ofpalpation aids. This study was designed to find outwhether video instruction tapes or palpation aids arelikely to be effective in improving the detection of breastlumps during self-examination.
Materials and methods
Participants
The study group comprised participants who hadresponded to a notice asking for volunteers with noprevious experience of BSE and/or (in the case of men)clinical breast examination. The notice had been postedin the Department of Medical Psychology and Sociol-ogy, Justus-Liebig University, Giessen.
Materials
Spherical tumor equivalents ranging from 0.8 to 3.0 cmin diameter were inserted into 20 of 24 different sizedsilicone breast forms (Amoena Corporation, Raubling,Germany). The tumor equivalents were inserted fromthe back of the form and could not be seen by the study
participants. The pressure applied during palpation wasmeasured on a scale, which was placed underneath thebreast models.
The palpation aids (Vital Aid) were gifted by theSentida Corporation (Preston, WA, USA). The VitalAid comprises a pad made of polyurethane sheets filledwith a small amount of lubricant. According to themanufacturers, the Vital Aid, when placed on the breast,reduces skin-on-skin friction and increases sensitivity tolumps during self-examination. (Today the Vital Aid ismarketed now as the LivAid.)
Instruction in BSE was provided by a short(4.45 min) video (Brust bewusst – Selbstabtasten mitFingerspitzengefuhl). This video is recommended byvarious German societies involved in the treatment ofbreast cancer and received the Intermedia-GlobeFinalist Diploma at the World Media Festival in 2004.It is available commercially and was obtained from theAktion Bewusstsein fur Brustkrebs eV (Schriesheim,Germany).
Study design
Participants were randomly assigned to experimentalgroups as follows: instruction video seen versus not seenand palpation aids used versus aids not used. A 2�2�2multifactorial study design incorporating these experi-mental interventions and gender was set up (Table 1).The dependent variable was the rate of accurate tumordetection, defined as the sum of tumors detected within20 s (mean sum of correct positive hits (CPH)). The timelimitation was set to ensure that differences in resultswere a result of different treatments only. Only thosesilicon breast forms with tumor detection rates of atleast 10% but no more than 90% were to be included inorder to avoid the possibility of distortion throughceiling or ground effects.
Psychological measures
In order to avoid any possible effect that personalityvariables such as anxiety might have on the number of
Table 1. Multifactorial study design
Gender
Male Female Total
Palpation aid
No Instruction video No 12 12 24
Yes 12 13 25
Total 24 25 49
Palpation aid
Yes Instruction video No 12 13 25
Yes 13 12 25
Total 25 25 50
Total 49 50 99
168 R von Georgi et al.
CPH registered, we used Carver and White’s BehavioralInhibition System and Behavioral Activation SystemScale (BIS/BAS) [12,13] and the Self-Construct Inven-tory to examine the study population [14]. The BIS/BASScale measures general anxiety (BIS) and impulsivenessand a positive approach (BAS). It is based on the theoryof Gray and McNaughton [12,13]. The Self-ConstructInventory comprises five basic dimensions of self: egostrength, attractiveness, confidence, orderliness andforcefulness. Participants also answered questionsdesigned to assess their perceptions of BSE and of theimportance of secondary preventive measures in breastcancer. The questionnaire was based upon earlier workon subjects’ knowledge of screening and detection [15].
Procedure
The breast examination procedures were conductedindividually in a specially prepared room in theDepartment of Medical Psychology and Sociology.After randomization to experimental groups, partici-pants completed the psychological and knowledgequestionnaire. All subjects were then shown how toexamine the breast forms (Figure 1), and were instructednot to apply a force of more than 80 N during palpa-tion. Those subjects randomized to video interventionthen watched the instruction video. Finally, subjectspalpated the breast models with or without a palpationaid. During that time, the investigator (FT) assessed thepalpation pressure they applied. After 20 s the partici-pant had to state whether he or she had found a tumorequivalent and where it was located in relation to themammilla. Only if the position and size of the tumorcorresponded with the real situation, was the test ratedcorrect. Thus, it was possible to test whether the par-ticipant really had palpated a tumor, or was mistaken.
Results
Participants
One hundred volunteers participated in the study – 50men and 50 women. Their mean (±SD) age was 25.6(±4.9) years (median 25 years). The mean (±SD) agefor men was 26.6 (±6.2) years (median 25 years), whilethat for women was 24.6 (±2.9) years (median24 years). Most volunteers were medical students, but 22were studying non-medical subjects and 3 were notuniversity students. None of the participants had pre-vious experience of BSE and/or clinical breast palpation.
We excluded the data from one participant before thefinal analysis. This was because the results of bothexperimental tests and psychological tests in this indi-vidual were outside the normal range.
Silicon breast forms
Results for 14 breast forms were included in the finalanalysis. Ten silicon models were excluded from themean CPH values as the tumor equivalents inserted intothem were either too easy or too difficult to detect andthis fact might have contributed to a ceiling/ground ef-fect.
Palpation pressure
Palpation pressure exceeded 80 N in one case only.
Statistical analyses of the study group
Validity of CPH valueThe mean (±SD) CPH value for all participants was0.57 (±0.17). We tested the proximity of the data set to
Figure 1. Testing for possible tumor equivalents. The picture shows one of the breast models during palpation as well as the evaluation protocol.
Can instruction videos and palpation aids improve BSE outcomes? 169
the normal distribution. Skewness and kurtosis valueswere S=)0.001 and K=)0.55 respectively. TheKolmogoroff–Smirnoff test for the entire group shows avalue of p=0.009 [0.105; DF=99]. The Shapiro–Wilkstest for the factors: sex, instruction video, or palpationaids, shows a p value of ‡0.026 [‡0.946; DF ‡ 49]. TheLevene test for equality of variance between treatmentsteps gave a p value of ‡0.96 [F £ 1.08; DF=98]. Theseresults mean that the CPH values were normally dis-tributed except for one small peak (Kolmogoroff–Smirnoff test and kurtosis of the CPH value). Thus, theCPH value was a valid indicator of the effectiveness ofbreast examination in this study.
Possible influence of personality variablesWe tested the effects of all personal variables on groupdifferences (2�2�2 ANOVA) and correlations. Apartfrom gender, no significant differences where observed(p‡1.0). We observed significant main effects in relationto the following traits: BIS/BAS sensitivity (women,mean (±SD) 19.46 (±3.21) versus men 15.26 (±3.77);p[DF=1; F=35.59] £ 0.001; r=0.52; p £ 0.001);Self Concept Inventory ego-strength (women 31.71(±7.93); versus men 40.35 (±8,19); p[DF=1;F=25,78] £ 0.001; r=)0.48; p £ 0.001) and confi-dence (women 34.7 (±7.04) versus men 38.57 (±5.91),p[DF=1; F=7,70]=0,007; r=)0.28; p=0.005). Thus,female participants in our study were more anxious andunsure and not as confident as the men. As a result, weanalyzed the uncorrected as well as the gender traitcorrected correlations between CPH values and thepersonal variables. These correlations showed no sig-nificant differences (r £ 0.104; p ‡ 0.30) and we there-fore assumed that the CPH values were not influencedby personal traits.
Role of instruction video and palpation aid
Mean CPHThe effects of the variables, video instruction and pal-pation aid on CPH values, and possible differences
related to gender were tested by 2�2�2 multifactorialANOVA for differences in the mean CPH values(Table 2). In addition, mean values were correctedaccording to the initial analysis for potential effects(COVAR). No influence on the CPH value was seen forgender, the instruction video or any interaction effects ofthe two. However, use of the palpation aid resulted in ahighly significant main effect in both ANOVA andCOVAR analyses (p=0.00003; Table 2). Participantswho used the palpation aid had a significantly lowerCPH value. This suggests that using the palpation aidresulted in greater difficulty in identifying the tumorequivalents (aid, mean (±SD) 0.52 (0.16) versus no aid,0.62 (±0.15)).
Regression analysisWomen using the palpation aid had a lower mean CPHvalue (0.48; Table 3). Although this interaction effectwas not statistically significant (gender�palpation aid;p=0.244; Table 3) there may have been regressiveinteraction effects. We therefore examined all the vari-ables assessed (gender, palpation aid, video instruction,all personality scales, age, and number of semestersstudied) by Classification and Regression Tree analysisusing SPSS AnswerTree (Breiman et al. 1984, SPSSAnswerTree V3.0, SPSS, Chicago, IL). The parent nodewas defined as 10 and the child node was 5. The mini-mum change of impurity was set to 0.001. This analysisshowed a distribution of the CPH value in the entiregroup only for the nominally scaled variable ‘palpationaid’ (improvement=0.0045). The estimated risk was0.022(SE=0.003), meaning that a false assignment tothe two groups based on the CPH value was only 2%.Thus, the lack of homogeneity of the CPH value in theentire group is predominantly a result of the use of thepalpation aid. The predicted mean (±SD) values were0.64(±0.15; n=49) with the palpation aid and0.50(±0.51; n=49) without it, and these are identical tovalues in Table 3. Thus, regression analysis showed thatthe CPH value was not influenced by any variable orinteraction effect except the use of the palpation aid.
Table 2. Effects of variables on mean correct positive hit values, analysis of variance (ANOVA) and covariance (COVAR)
ANOVA COVAR
F(DF=1) p p
Gender 0.064 0.800 0.753
Palpation aid 19.019 >0.001 >0.001
Instruction video 0.052 0.820 0.708
Gender�palpation aid 1.375 0.244 0.341
Gender�instruction video 0.597 0.442 0.432
Palpation aid�instruction video 0.345 0.559 0.509
Gender�palpation aid�instruction video 1.934 0.168 0.095
COVAR: Results of the analysis of variance corrected by the following psychological covariates: Behavioral Inhibition System and Behavioral
Activation System Scale (BIS/BAS): p(DF=1; F=0.902)=0.345; ego strength (Self-Construct Inventory) p(DF=1; F=1.102)=0.297; confidence
(Self-Construct Inventory) p(DF=1; F=0.030)=0.862.
170 R von Georgi et al.
Discussion
This study shows that the effectiveness of detectingtumor equivalents in an experimental setting was notimproved by video instruction and was not influencedby other factors such as gender and personality traits.Interestingly, and contrasting with earlier expectations,palpation aids were associated with poorer detectionrates. However, the validity of our conclusion on theeffectiveness of the palpation aid is limited by the factthat during BSE women receive neural impulses fromboth their fingers and the breast. The palpation aidwill reduce the neural information because the frictionbetween the skin of the breast and the fingers is reduced.These conditions could not be simulated in this experi-mental setting.
Earlier studies on BSE assessed various aspects of themethod. The acceptance of the procedure was found tobe determined by the following: anxiety about cancer,self-consciousness, perceived breast cancer risk, infor-mation on the procedure, and belief in its efficacy[16–24]. It was also reported that instruction andtraining in BSE are well perceived by most participantsand improve palpation depth and time of palpation[25,26]. However, few trials have investigated the effec-tiveness of the procedure. One trial focused on themethod of self-examination. It compared detection ratesin subjects who used linear palpation with those insubjects using palpation in concentric circles, but foundno differences [27]. One study suggests that training inself-examination techniques may have some effect. Itshowed that after training, women were as accurate asphysicians in finding tumor equivalents, and this isbelieved to improve, at least modestly, the early detec-tion of breast cancer [28,29]. Unfortunately, no pub-lished study has evaluated detection rates. Anunpublished study by Sonoo et al, which evaluated theuse of palpation aids, showed some advantage for these.However, the study data available leave many questionsunanswered, in particular, the number of lumps, whichwere examined histologically.
In conclusion, our results clearly show that tech-niques which claim to improve the effectiveness of BSE –
the use of instruction videos and palpation aids – do nothold up under critical analysis. When our results areadded to those from large randomized studies of BSE,providing level 1 evidence of its lack of effectiveness, it isdifficult to understand why the procedure continues tobe promoted by various groups [30]. To date, there is noevidence supporting the clinical use of BSE. Advocatesof the procedure should be aware of its shortcomingsand exercise caution in recommending additionalscreening methods such as mammography and ultraso-nography. Reasons behind the continuing recommen-dation of BSE which have not yet been assessedscientifically may include a failure to accept that theearly detection of breast cancer cannot be improved byself-examination despite existing evidence and a desireto make women fully aware of the problem. Counselorsgiving advice on BSE also include information on breastcancer and on the possibilities for its cure. This infor-mation may provide women with hope and reduce thelarge number (66%) recorded in recent studies whodelay seeking help and present with advanced breastcancer [31].
Acknowledgements
The authors are very grateful to the Amoena Corpora-tion (Raubling, Germany), especially Mr Wild, forproviding the breast forms and to the Sentida Corpo-ration (Preston, WA) for supplying the palpation aids.
References
1. Vahabi M: Breast cancer screening methods: a review of the evi-
dence. Health Care Women Int 24: 773–793, 2003
2. Harvey BJ, Miller AB, Baines CJ, Corey PN: Effect of breast self-
examination techniques on the risk of death from breast cancer.
CMAJ 157: 1205–1212, 1997
3. Coates RJ, Ushler RJ, Brogan DJ, Gammon MD, Malone KE,
Swanson CA, Flagg EW, Brinton LA: Patterns and predictors
of the breast cancer detection methods in women under 45 years
of age (United States). Cancer Causes Control 12: 431–442,
2001
Table 3. Correct positive hit values (mean (±SD)) in relation to intervention and gender
Gender
Male Female Sum
Palpation aid
No Instruction video No 0.60±0.15 0.70±0.15 0.65±0.15
Yes 0.64±0.15 0.60±0.15 0.62±0.15
Sum 0.62±0.15 0.65±0.15 0.64±0.15
Palpation aid
Yes Instruction video No 0.53±0.15 0.46±0.16 0.50±0.15
Yes 0.52±0.17 0.49±0.15 0.51±0.16
Sum 0.52±0.16 0.48±0.15 0.50±0.15
Total 0.57±0.15 0.56±0.15 0.57±0.17
Can instruction videos and palpation aids improve BSE outcomes? 171
4. Koster JP, Gotzsche PC: Regular self-examination or clinical
examinations for early detection of breast cancer. Cochrane
Database Syst Rev (2): CD003373, 2003
5. Hackshaw AK, Paul EA: Breast self-examination and death from
breast cancer: a meta-analysis. Br J Cancer 88: 1047–1053, 2003
6. Thomas DB, Gao DL, Self SG, Allison CJ, Tao Y, Mahloch J,
Ray R, Qin Q, Presley R, Porter P: Randomized trial of breast
self-examination in Shanghai: methodology and preliminary
results. J Natl Cancer Inst 89: 355–365, 1997
7. Thomas DB, Gao DL, Ray R, Wang WW, Allison CJ, Chen FL,
Porter P, Hu YW, Zhao GL, Pan LD, Li W, Wu C, Coriaty Z,
Evans I, Lin MG, Stalsberg H, Self SG: Randomized trial of
breast self-examination in Shanghai: final results. J Natl Cancer
Inst 94: 1445–1457, 2002
8. Semiglazov VF, Moiseyenko VM, Bavli Migmanova NSh JL,
Seleznyov NK, Poppva RT, Ivanova OA, Orlov AA, Chagunava
OA, Barash NJ: The role of breast self-examination in early breast
cancer detection (results of the 5-years USSR/WHO randomized
study in Leningrad). Eur J Epidemiol 8: 498–502, 1992
9. Holmberg L, Ekbom A, Calle E, Mokdad A, Byers T: Breast
cancer mortality in relation to self-reported use of breast self-
examination. Breast Cancer Res Treat 43: 137–140, 1997
10. Newcomb PA, Weiss NS, Storer BE, Scholes D, Young BE, Voigt
LF: Breast self-examination in relation to the occurrence of
advanced breast cancer. J Natl Cancer Inst 83: 260–265, 1991
11. Green BB, Taplin SH: Breast cancer screening controversies. J Am
Board Fam Pract 16: 233–241, 2003
12. Carver CS, White TL: Behavioral inhibition, behavioral activa-
tion, and affective responses to impending reward and punishment:
the BIS/BAS scales. J Pers Soc Psychol 67: 319–333, 1994
13. Stroebel A., Beaducel Debner Brocke A. S. B: Eine deutschsp-
rachige Version des BIS/BAS-Fragebogens von Carver und White.
Zeitschr Different Diagn Psychol 22: 216–227, 2001
14. von Georgi R, Beckmann D: Selbstkonzept-Inventar. Verlag Hans
Huber Bern, Gottingen, Toronto, Seattle, 2004
15. McCance KL, Mooney KH, Smith KR, Field R: Validity and
reliability of a breast cancer knowledge test. Am J Prev Med 6: 93–
98, 1990
16. Janda M, Obermair A, Haidinger G, Waldhoer T, Vutuc C:
Austrian women’s attitudes towards and knowledge of breast self-
examination. J Cancer Educ 15: 91–94, 2000
17. Janda M, Stanek C, Newman B, Obermair A, Trimmel M: Impact
of videotaped information on frequency and confidence of breast
self-examination. Breast Cancer Res Treat 73: 37–43, 2002
18. Rutledge DN, Barsevick A, Knobf MT, Bookbinder M: Breast
cancer detection: knowledge, attitudes, and behaviors of women
from Pennsylvania. Oncol Nurs Forum 28: 1032–1040, 2001
19. Erblich J, Bovbjerg DH, Valdimarsdottir HB: Psychological
distress, health beliefs, and frequency of breast self-examination.
J Behav Med 23: 277–292, 2000
20. Wood RY, Duffy ME, Morris SJ, Carnes JE: The effect of an
educational intervention on promoting breast self-examination in
older African American and Caucasian women. Oncol Nurs
Forum 29: 1081–1090, 2002
21. Yelland MJ, Rice DE, Ward AE, Bain C, Siskind V, Schofield F:
A profile of Australian women practicing breast self-examination.
Asia Pac J Public Health 5: 307–312, 1991
22. Olson RL, Mitchell ES: Self-confidence as a critical factor in breast
self-examination. J Obstet Gynecol Neonatal Nurs 18: 476–481,
1989
23. Nettles-Carlson B, Field ML, Friedman BJ, Smith LS: Effective-
ness of teaching breast self examination during office visits. Res
Nurs Health 11: 41–50, 1988
24. Champion VL: Effect of knowledge, teaching method, confidence
and social influence on breast self-examination behavior. Image J
Nurs Sch 21: 76–80, 1989
25. Stefanek ME, Wilcox P, Huelskamp AM: Breast self-examination
proficiency and training effects: women at increased risk of breast
cancer. Cancer Epidemiol Biomarkers Prev 1: 591–596, 1992
26. Leight SB, Leslie NS: Development of a competency-based cur-
riculum for training women in breast self-examination skills. J Am
Acad Nurse Pract 10: 297–302, 1998
27. Murali ME, Crabtree K: Comparison of two breast self-exami-
nation palpation techniques. Cancer Nurs 15: 276–282, 1992
28. Jacob TC, Penn NE, Giebink J, Bastien R: A comparison of breast
self-examination and clinical examination. J Natl Med Assoc 86:
40–4, 1994
29. Bobo JK, Lee NC, Thames SF: Findings from 752,081 clinical
breast examinations reported to a national screening program
from 1995 through 1998. J Natl Cancer Inst 92: 971–976, 2000
30. Anderson BO, Braun S, Lim S, Smith RA, Taplin S: Thomas DB
for the Global Summit Early Detection panel: early detection of
breast cancer in countries with limited resources. Breast J 9(Suppl.
2): S51–S59, 2003
31. Harirchi I, Ghaemmaghami F, Karbakhsh M, Moghimi R,
Mazaherie H.: Patient delay in women presenting with advanced
breast cancer: an Iranian study. Public Health 119: 885–891, 2005
Address for offprints and correspondence: Karsten Munstedt, Univer-
sitatsfrauenklinik Giessen, Klinikstrasse 32, 35392, Giessen,
Germany; Tel.: +49-641-99-45200; Fax: +49-641-99-45139; E-mail:
172 R von Georgi et al.
