aUniversity ofCarolina.bDepartment ofof Medicine, DcRiverside RadiodSeattle RadioloeDepartment oNorth Carolina.

ª 2014 America

1546-1440/14/$

Predictors of Pain Experiencedby Women During PercutaneousImaging-Guided Breast BiopsiesAdrianne E. Soo, BSa, Rebecca A. Shelby, PhDb, Lauren S. Miller, MDc,

Melissa Hayes Balmadrid, MDd, Karen S. Johnson, MDe, Anava A. Wren, MSb,

Sora C. Yoon, MDe, Francis J. Keefe, PhDb, Mary Scott Soo, MDe

Purpose: The purpose of this study was to evaluate pain experienced during imaging-guided core-needlebreast biopsies and to identify factors that predict increased pain perception during procedures.

Methods: In this institutional review boardeapproved, HIPAA-compliant protocol, 136 women under-going stereotactically or ultrasound-guided breast biopsy or cyst aspiration were recruited and providedwritten informed consent. Participants filled out questionnaires assessing anticipated biopsy pain, ongoingbreast pain, pain experienced during biopsy, catastrophic thoughts about pain during biopsy, anxiety,perceived communication with the radiologist, chronic life stress, and demographic and medical information.Procedure type, experience level of the radiologist performing the biopsy, number of biopsies, breast density,histology, and tumor size were recorded for each patient. Data were analyzed using Spearman’s r correlationsand a probit regression model.

Results: No pain (0 out of 10) was reported by 39.7% of women, mild pain (1e3 out of 10) by 48.5%, andmoderate to severe pain (�4 out of 10) by 11.8% (n ¼ 16). Significant (P < .05) predictors of greater biopsypain in the probit regression model included younger age, greater prebiopsy breast pain, higher anticipatedbiopsy pain, and undergoing a stereotactic procedure. Anticipated biopsy pain correlated most strongly withbiopsy pain (b ¼ .27, P ¼ .004).

Conclusions: Most patients report minimal pain during imaging-guided biopsy procedures. Womenexperiencing greater pain levels tended to report higher anticipated pain before the procedure. Communi-cation with patients before biopsy regarding minimal average pain reported during biopsy and encouragementto make use of coping strategies may reduce patient anxiety and anticipated pain.

Key Words: Imaging-guided breast biopsy, breast pain, anticipated pain

J Am Coll Radiol 2014;11:709-716. Copyright © 2014 American College of Radiology

INTRODUCTIONPain experienced by women during percutaneousimaging-guided breast biopsies is an important factor inthe overall biopsy experience. Evaluating the degree ofpain is important for ensuring that pain managementtechniques are appropriate and effective, providing pa-tients with reasonable expectations about the biopsyexperience, and may influence adherence to future

North Carolina School of Medicine, Chapel Hill, North

Psychiatry and Behavioral Sciences, Duke University Schoolurham, North Carolina.

logy and Interventional Associates, Columbus, Ohio.

gists, Seattle, Washington.

f Radiology, Duke University Medical Center, Durham,

n College of Radiology

36.00 � http://dx.doi.org/10.1016/j.jacr.2014.01.013

mammographic screening [1]. In addition, effective painmanagement has potential financial implications. CMSrecently finalized the details of a new reimbursementplan that adjusts payments on the basis of patientsatisfaction [2]. The Hospital Consumer Assessment ofHealthcare Providers and Systems survey used by CMSto evaluate patient satisfaction includes a questionregarding pain management [2], suggesting that effective

Corresponding author and reprints: Mary Scott Soo, MD, Duke UniversityMedical Center, Department of Radiology, Box 3808, Room 24244B, 2ndFloor Red Zone, Duke South Clinic, Trent Drive, Durham, NC 27710;e-mail: soo00002@mc.duke.edu.

This study was supported in part by a pilot grant from The John Tem-pleton Foundation (grant 12111) through the Center for Spirituality, The-ology and Health at the Duke University Medical Center (Durham, NorthCarolina).

709

710 Journal of the American College of Radiology/Vol. 11 No. 7 July 2014

pain management is considered an important compo-nent of patient care that could eventually affect practicerevenues.Previous studies reported average pain scores during

imaging-guided breast biopsy ranging from 2.0 to 5.8out of 10 for multipass core-needle biopsies and 3.3 to4.6 out of 10 for vacuum-assisted (VA) biopsies, on thebasis of visual analog or fixed-interval rating scales givenshortly after the procedure [3-10]. Although these meanpain levels are low to moderate, higher levels of painhave been identified among certain patient subsets.Research has found greater pain and discomfort tocorrelate with greater prebiopsy anxiety, biopsies per-formed during the luteal phase of the menstrual cycle,deeper lesions, longer procedures, greater number ofprocedures, and larger needle diameter [3,4,7-9]. Severalstudies have looked at the relationship between meno-pausal status and pain [4,10], finding no significantcorrelation.There have been conflicting results regarding other

pain predictors. Whereas one study showed that VAdevices were more painful than multipass devices [3],another showed the reverse trend [6], and still anothershowed no correlation [9]. Similar discrepancies werefound regarding the significance of lesion histology,patient age, breast density, and biopsy operator in pre-dicting biopsy pain [4,7-10].There have been no prior studies evaluating the

impact of radiologist-patient communication, antici-pated biopsy pain, or pain catastrophizing (ie, the ten-dency to focus on and exaggerate the threat value ofpainful stimuli and negatively evaluate one’s own abilityto deal with pain) on biopsy-related pain [11]. Higheranticipated pain, greater pain catastrophizing, and lowerphysician-patient communication have all been corre-lated with higher pain scores during or after othermedical interventions [12-15], suggesting that thesefactors may affect pain experienced during breast biopsy.Given the impact of pain on the overall biopsy

experience, the purposes of this study were to (1) eval-uate pain experienced by women during imaging-guidedcore-needle breast biopsy and cyst aspiration and (2)identify factors that predict increased pain perceptionduring these procedures. Identifying patients prone toexperiencing pain during procedures could facilitateimplementation of directed pain-reducing interventions,improving the overall breast biopsy experience.

METHODS

ParticipantsFrom August 2010 through February 2011, 207 of 818women at our breast center undergoing ultrasound-guided or stereotactically guided core-needle breast bi-opsy or ultrasound-guided diagnostic cyst aspirationwere invited to participate in this prospective pilot studyon the days of their procedures. Women were invited on

the basis of the following inclusion criteria: (1) aged�21 years, (2) presented for a percutaneous imaging-guided diagnostic procedure, (3) were able to speakand read English, (4) were able to provide writteninformed consent, and (5) had not undergone imaging-guided breast biopsy in the previous 6 months. Thestudy was HIPAA compliant and performed withinstitutional review board approval. Data from thispatient population examining the impact of radiologist-patient communication on anxiety have been previouslypublished [16], but that publication did not report anyof the biopsy pain data herein.

Of 152 women providing informed consent,16 participants were then excluded because 4 did notundergo biopsies, 1 had cognitive impairment inter-fering with questionnaire completion, and 11 did notcomplete the study measures, resulting in 136 women inthis sample.

ProceduresAs part of a larger ongoing study of patient adherence tomammography after biopsy, study participantscompleted written questionnaires immediately beforeand after biopsy, requiring approximately 10 min and20 to 30 min, respectively. Of 284 questions, 113pertained to this investigation of biopsy pain; prebiopsyquestionnaires assessed ongoing breast pain, anticipatorybiopsy pain, and prebiopsy anxiety. Postbiopsy ques-tionnaires evaluated pain during biopsy, catastrophicthoughts about pain during biopsy, postbiopsy anxiety,and communication with the biopsy radiologist; socio-demographic and medical characteristics were alsoassessed.

In 82 women, ultrasound-guided core-needle biopsieswere performed using 14-gauge (n ¼ 67) or 18-gauge(n ¼ 15) multipass devices (Achieve; Cardinal Health,Dublin, Ohio). Forty-one patients underwent stereo-tactic biopsies using 9-gauge VA probes (EVIVA; Suros,Indianapolis, Indiana) on a prone stereotactic table(Multicare Platinum; Lorad, Danbury, Connecticut).Thirteen women underwent cyst aspiration alone, using18-gauge (n ¼ 4), 19-gauge (n ¼ 1), 20-gauge (n ¼ 1),21-gauge (n ¼ 5), or 25-gauge (n ¼ 2) needles. Twopatients underwent both cyst aspiration and core biopsy.

During biopsy recommendation, patients receivedwritten and verbal biopsy information with guidelines toavoid aspirin and nonsteroidal anti-inflammatory agents.Seven breast imaging radiologists performed all pro-cedures, with or without the assistance of residents orfellows. Local anesthetic was administered using thefollowing general approach: <5 mL of 1% lidocaine wasinjected intradermally and subcutaneously using a25-gauge needle, followed by injection of approximately3 to 10 mL of lidocaine with epinephrine within andaround the lesion using a 22-gauge needle. After theprocedure, the biopsy site was compressed to achievehemostasis. The procedure type, number of radiologists

Soo et al/Predictors of Pain During Percutaneous Imaging-Guided Breast Biopsies 711

and their experience levels (ie, attending radiologistalone vs attending radiologist with resident or fellow),number of biopsy sites, needle gauge, breast density,histology (invasive carcinoma, ductal carcinoma in situ[DCIS], atypical lesions, or benign), and size of atypicalor malignant lesions were recorded.

Measures

Ongoing Breast Pain. Two questions from the BriefPain Inventory addressing average pain in the pastmonth (eg, “In the past month, on the AVERAGE, howintense was your breast pain?”) and present pain (eg,“How would you rate your breast pain RIGHTNOW?”) [17] were averaged, creating a total pain scoreassessing ongoing breast pain intensity on a scale rangingfrom 0 to 10 (0 ¼ no pain, 10 ¼ pain as bad as it couldbe; correlation between items, r ¼ 0.81).

Anticipatory Biopsy Pain. One item evaluated antici-pated pain during biopsy, asking “How much pain doyou think you will experience during your biopsytoday?” on a scale ranging from 0 to 10 (0 ¼ no pain,10 ¼ pain as bad as it could be).

Pain Experienced During Biopsy. Pain during biopsywas measured using 4 items (rated 0 ¼ no pain to 10 ¼pain as bad as it could be) drawn from the Brief PainInventory’s Pain Intensity Scale [17]. The items assessedworst, least, average, and present pain. The items wereaveraged, creating a total pain score (Cronbach’sa ¼ .84).

Pain Catastrophizing. Catastrophic thoughts aboutpain during biopsy were evaluated using two items fromthe Coping Strategies Questionnaire Short-Form (eg, “Itwas terrible and I thought it was never going to get anybetter” and “I felt I couldn’t stand it any more”) [14].Instructions were modified, asking about thoughtsoccurring specifically during biopsy. Items were aver-aged, giving a total score on a scale ranging from 0 to 6(0 ¼ not at all, 6 ¼ a lot; correlation between items,r ¼ 0.78).

Anxiety. The 20-item self-report State Anxiety Scale ofthe State-Trait Anxiety Inventory was used to assessprebiopsy and postbiopsy anxiety on a scale of 1 (not atall) to 4 (very much) [18]. Items were summed, creatinga total score (possible range, 20e80), with higher scoresindicating higher anxiety. Internal consistency was high(Cronbach’s a ¼ .91e.92).

Chronic Life Stress. Chronic life stress was measuredusing a 51-item questionnaire assessing general orambient problems, financial issues, work, marriage andrelationship issues, parental concerns, family, social life,residence, and health issues [19]. Responses include “nottrue” (0), “somewhat true” (1), and “very true” (2) andwere summed, creating a total score (potential range,0e102), with higher scores indicating greater chronic

life stress. Internal consistency was high (Cronbach’sa ¼ .91).

Perceived Communication. Patients’ perceptions ofcommunication with the biopsy radiologist weremeasured using a modified version of the Questionnaireon the Quality of Physician-Patient Interaction [20].Example items include “The radiologist’s explanationswere easy to understand” and “The radiologist gave meenough time to ask questions.” One item not pertinentto the biopsy scenario was dropped, and wording ofthe remaining 13 items was revised, using the termradiologist instead of physician. Items were rated ona scale ranging from 1 to 5 scale (1 ¼ do not agree,5 ¼ strongly agree) and summed, creating a total score(possible range, 13e65) with higher scores indicatingbetter perceived communication. Internal consistencywas high (Cronbach’s a ¼ .93e.95).

Demographic and Medical Information. Patientscompleted questionnaires assessing age, race, menopausalstatus, education, marital status, history of depression oranxiety, cancer history, and history of breast biopsy and/orsurgery.

Statistical AnalysisDescriptive statistics were computed for patient de-mographic characteristics, medical variables, and studyvariables. Several variables had non-normal distributions(ie, biopsy pain, ongoing breast pain, and pain cata-strophizing); therefore, bivariate analyses were con-ducted using Spearman’s r and Kruskal-Wallis tests toexamine relationships between study variables and bi-opsy pain. Variables significantly associated (P < .05)with biopsy pain were then used as predictors in a probitregression model. Because of its skewed distribution,pain was coded as none (0), mild (1e3), or moderate tosevere (4e10) for the regression analysis. Mplus version5.1 (Muthén & Muthén, Los Angeles, California) wasused to produce parameter estimates using a robustweighted least squares estimator.

RESULTS

Descriptive StatisticsDescriptive statistics are displayed in Table 1. Partici-pants (n ¼ 136) averaged 52 � 3 years of age and15.5 � 3.3 years of education and were predominantlyCaucasian (71%) and married (58%). Forty-sevenpercent of women had undergone previous breast bi-opsies, 35% had undergone previous breast surgery, and19% had personal histories of breast cancer. For pain-related variables (0e10 scale), the mean breast painbefore biopsy was 0.91 � 1.62 (median, 0; range,0e7.5), the mean anticipated biopsy pain was 4.4 �2.73 (median, 4; range, 0e10), and the mean painduring biopsy was 1.25 � 1.42 (median, 0.75; range,0e6.5). Figure 1 displays the distribution of biopsy

Table 1. Sample description and descriptive statisticsVariable % (n) Mean – SD Median Range

Age (y) 52.13 � 13.00 62.00 32e65Education (y) 15.54 � 3.31 16.00 7e25RaceCaucasian 71.3 (97)African American/black 25.7 (35)Asian 2.2 (3)Native American 0.7 (1)

Married 58.1 (79)Previous breast biopsy 47.1 (64)Previous breast surgery 35.3 (48)Personal breast cancer history 19.1 (26)First-degree relative with breast cancer 21.3 (29)Postmenopausal 50.0 (68)Comorbid illness (�1 illness) 87.9 (106)Body mass index (kg/m2) 28.17 � 7.21 27.00 17.6e53.5Pain medication taken day of biopsy 24.0 (32)Breast densityPredominantly fatty 7.9 (10)Scattered fibroglandular 27.2 (37)Heterogeneously dense 43.4 (59)Extremely dense 14.7 (20)Unknown 7.4 (10)

Type of biopsyCyst aspiration only 9.6 (13)Ultrasound-guided core needle 60.3 (82)Stereotactically guided core needle 30.1 (41)

Number of biopsy procedure1 83.1 (113)2 11.8 (16)3 5.1 (7)

Biopsy resultBenign 66.9 (91)Atypia 5.9 (8)Ductal carcinoma in situ 3.7 (5)Invasive breast cancer 23.5 (32)

Biopsy pain 1.25 � 1.42 0.75 0e6.5Ongoing breast pain before biopsy 0.91 � 1.62 0.00 0e7.5Anticipated biopsy pain 4.43 � 2.73 4.00 0e10Pain catastrophizing 0.36 � 1.06 0.00 0e6Anxiety 44.29 � 12.30 43.50 20e77Chronic life stress 15.99 � 11.67 12.50 0e69Communication with radiologist 58.11 � 8.24 62.00 32e65Radiologist years of experience 10.86 � 5.95 12.00 1e18

712 Journal of the American College of Radiology/Vol. 11 No. 7 July 2014

pain. No pain (0) was reported by 54 women (39.7%),mild pain (1e3) by 66 women (48.5%), and moderateto severe pain (�4) by 16 women (11.8%).

CorrelationsCorrelates of perceived pain during biopsy were exam-ined using Spearman’s r (Table 2). Older women re-ported less pain during biopsy (r ¼ �0.23, P ¼ .007).Having denser breast tissue (r ¼ 0.25, P ¼ .004), agreater number of procedures (r ¼ 0.29, P ¼ .001),larger needle size (ie, smaller needle gauge; r ¼ �0.28,P ¼ .001), and stereotactic procedures (r ¼ 0.27,P¼ .001) were significantly associatedwith greater biopsypain. Greater pain during biopsy was also associated withgreater anxiety before biopsy (r¼ 0.20, P¼ .02), higherbreast pain before biopsy (r ¼ 0.37, P < .001), greater

anticipated biopsy pain (r ¼ 0.42, P < .001), greaterpain catastrophizing (r ¼ 0.43, P < .001), and highertotal life stress (r ¼ 0.32, P < .001). No significantcorrelations were found between biopsy pain and bodymass index, number of comorbid illnesses, previousbreast biopsy or breast surgery, having a personalhistory of breast cancer or a first-degree relative withbreast cancer, years of education, use of oral painmedications on the day of biopsy, menopausal status,perceived communication with the radiologist, or radi-ologist’s years of experience.

Levels of pain during biopsy are displayed in Table 3by type of biopsy procedure and biopsy result. Painsignificantly differed by procedure type (Kruskal-Wallistest, c2 ¼ 11.86, df ¼ 2, P ¼ .003). Post hoc testsusing Bonferroni’s correction indicated that women

Table 2. Correlations (Spearman’s r) with biopsy pain

VariableCorrelationCoefficient P

Demographic characteristicsAge L.23 .007Education �.05 .57Race* �.10 .23Marital status† �.07 .43

Medical characteristicsPrevious breast biopsy .07 .45Previous breast surgery �.03 .75Personal breast cancer history .007 .93First-degree relative with breast cancer .001 .99Postmenopausal �.13 .14Body mass index �.07 .45Number of comorbid illnesses �.11 .23Pain medication taken day of biopsy �.05 .58Breast density .25 .004

Biopsy-related variablesType of biopsyz .27 .001Number of procedures .29 .001Needle gauge L.28 .001Biopsy resultx .15 .08

Radiologist-related variablesCommunication with radiologist �.08 .36Radiologist years of experience �.05 .60

Pain and psychosocial variablesOngoing breast pain before biopsy .37 <.001Anticipated biopsy pain .42 <.001Pain catastrophizing .43 <.001Anxiety .20 .02Chronic life stress .32 <.001

Note: Significant correlations (P < .05) are in boldface type.*Coded as 0 ¼ African American, Asian, or Native American,1 ¼ white/Caucasian.†Coded as 0 ¼ not married, 1 ¼ married.zCoded as 0 ¼ ultrasound-guided cyst aspiration or core-needle biopsy, 1 ¼ stereotactic biopsy.xCoded as 0 ¼ benign, 1 ¼ atypia, DCIS, or invasive breastcancer.

Fig 1. Distribution of pain during biopsy. The total painscore reflected the average of worst, least, average, andpresent pain reported by patients on a scale ranging from0 to 10.

Soo et al/Predictors of Pain During Percutaneous Imaging-Guided Breast Biopsies 713

undergoing stereotactic procedures reported significantlymore pain than women undergoing cyst aspiration orultrasound-guided biopsy. Although the correlation be-tween biopsy pain and subsequent benign versusabnormal (atypia, DCIS, or invasive cancer) biopsyresult was not significant, we further examined therelationship between pain and histology among womenreceiving abnormal results. There was no significantdifference in biopsy pain by specific histology (atypia vsDCIS vs invasive cancer; Kruskal-Wallis test, c2 ¼ 3.60,df ¼ 2, P ¼ .17). However, these results should beinterpreted with caution given the small numbers ofwomen with atypia (n ¼ 8) and DCIS (n ¼ 5).

Regression AnalysisProbit regression analysis was conducted to test theability of variables identified in bivariate analyses topredict pain during biopsy in multivariate analyses.Eight variables were included in the probit regressionmodel: biopsy type (coded as 0 ¼ ultrasound-guidedcyst aspiration or core-needle biopsy, 1 ¼ stereotactic-guided biopsy), number of procedures, age, ongoingbreast pain before biopsy, anticipated biopsy pain, paincatastrophizing, anxiety, and chronic life stress. Breastdensity and needle gauge were excluded from theregression model because breast density was confoundedby age (younger women had denser breast tissue;r¼�0.54, P< .001) and needle gauge was confoundedby procedure type (ie, all stereotactic procedures used9-gauge VA probes). Biopsy pain was coded as a 3-levelordinal variable: none (0), mild (1e3), or moderate tosevere (4e10). The probit regression analysis demon-strated that biopsy type, age, anticipated biopsy pain, andbreast pain before biopsy were significantly related tobiopsy pain (Table 4). Of these variables, anticipatedbiopsy pain demonstrated the strongest relationshipwith biopsy pain.

DISCUSSIONEvaluating pain associated with breast biopsy anddetermining predictors of pain are important steps forimproving women’s breast biopsy experiences. Ourstudy found overall low levels of pain (1.2 out of 10)reported by women during imaging-guided core-needlebreast biopsy and cyst aspiration. Approximately 40% ofwomen reported no pain during the procedure; amongwomen reporting pain, the average level was 2.3 � 1.4.The average biopsy pain was only slightly higher thanprebiopsy ongoing breast pain, suggesting that the bi-opsy procedure itself may account for only part of re-ported biopsy pain. In fact, the highest reported painduring biopsy (6.5 out of 10) was lower than the highestlevel of breast pain reported before the biopsy (7.5 outof 10).

Low levels of pain reported during biopsy support theefficacy of locally administered anesthesia, which shouldbe standard care for any percutaneous breast biopsyprocedure and may have even temporarily alleviated

Table 3. Biopsy pain by type of procedure and tumortype

VariableNo Pain

(0)Mild Pain(1e3)

Moderate toSevere Pain

(‡4)Type of biopsyCyst aspiration

(n ¼ 13)61.5% 30.8% 7.7%

Ultrasound-guidedcore needle(n ¼ 82)

43.9% 46.3% 9.8%

Stereotacticallyguided coreneedle (n ¼ 41)

26.5% 58.5% 17.1%

Biopsy resultBenign (n ¼ 91) 44.0% 46.2% 9.9%Atypia (n ¼ 8) 37.5% 62.5% 0.0%Ductal carcinoma

in situ (n ¼ 5)20.0% 20.0% 60.0%

Invasive breastcancer (n ¼ 32)

31.2% 56.2% 12.5%

714 Journal of the American College of Radiology/Vol. 11 No. 7 July 2014

prebiopsy breast pain in some cases. However, previousstudies using similar techniques for administering localanesthesia, needle gauge, biopsy procedure type, andpain measurement reported much higher levels of painfor both needle core (2.0e5.8 out of 10) and VA(3.3e4.61 out of 10) breast biopsies [3-10]. Althoughour study included cyst aspirations, which were slightlybut significantly less painful than other procedures, ourpain scores for stereotactically and ultrasound-guidedbiopsies were still lower than pain scores previously re-ported. Differences in biopsy device or technique, pa-tient demographics, or provider-related characteristicssuch as patient-radiologist interaction may havecontributed to the discrepancy.For procedure-related variables, lower needle gauge

and stereotactic procedures were associated with greaterpain. Because all stereotactic biopsies used VA deviceswith larger 9-gauge needles, compared with ultrasoundprocedures using multipass devices with smaller,14-gauge or 18-gauge needles, we cannot say definitivelywhether needle gauge alone, biopsy device (VA vsmultipass device), other differences between the pro-cedures (eg, prone vs supine positioning), or some

Table 4. Probit regression analysis of predictors of biopsy p

Variableb (Standardized

Coefficient)B

Type of biopsy* .18Number of procedures .17Age L.20Anxiety .01Ongoing breast pain before biopsy .17Anticipated biopsy pain .27Pain catastrophizing .18Chronic life stress .13

Note: Significant coefficients (P < .05) are in boldface type.*Coded as 0 ¼ cyst aspiration or ultrasound-guided core-needle bioperror.

combination of factors were responsible for the differ-ences in pain scores. We suspect that larger needle size,less comfortable prone position, and breast compressionof stereotactic procedures contributed to increased pain.Other patient-related characteristics associated withhigher biopsy pain included younger age, greater breastdensity, and greater ongoing breast pain, all found to besignificant predictors in the regression analysis. Aware-ness of these findings may help radiologists bettermanage patients’ pain during procedures. Histology,tumor size, radiologist’s experience level, menopausalstatus, and number of radiologists performing the biopsywere not associated with pain.

In addition to procedure and patient demographiccharacteristics, our bivariate analysis indicated relation-ships between patients’ psychological factors and biopsypain. Higher pain scores were correlated with higheranticipated biopsy pain, prebiopsy anxiety, and morecatastrophic thoughts about pain during biopsy. Whenall psychological variables were simultaneously examinedin the regression model, only anticipated biopsy painremained significant, suggesting the potential benefit ofinterventions targeted to decrease anticipatory biopsypain. However, efforts to reduce anxiety and cata-strophic thoughts may also be important, as these vari-ables were correlated with anticipatory biopsy pain andbiopsy pain (r ¼ 0.20e0.32, P < .05).

Although our observational study did not test in-terventions to decrease pain, existing techniques mayhelp, some warranting further study. Premedicationwith mild anxiolytic agents has been shown to reducediscomfort and anxiety during breast biopsy [21,22].Most practices, including ours, find their use unnec-essary, however, because they create logistic problemssuch as restricting patients from driving, therebyrequiring accompanying responsible adults, and limitingpatients’ ability to return to work that day [23].Methods such as relaxation or distraction techniquesmight help reduce pain and catastrophic thoughtswithout creating the logistic concerns. Previous studieshave shown that these techniques (eg, meditation,listening to music) and the use of self-calming or copingstatements can decrease perceived acute pain [24-26],

ain(UnstandardizedCoefficient) SE P

95% ConfidenceInterval for B

.56 .26 .03 .05 to 1.06

.46 .30 .13 �.13 to 1.05L.02 .01 .02 �.04 to �.003.001 .01 .92 �.02 to .02.15 .07 .03 .01 to .28.14 .05 .004 .05 to .23.25 .16 .11 �.06 to .55.02 .009 .07 �.002 to .03

sy, 1 ¼ stereotactically guided core needle biopsy. SE ¼ standard

Soo et al/Predictors of Pain During Percutaneous Imaging-Guided Breast Biopsies 715

and self-hypnosis has proved beneficial in decreasingbreast biopsy pain [23]. Even simple conversation be-tween the radiologist or technologist and the patientseems effective in our practice for monitoring anddiverting catastrophic thoughts during the procedure.Although there are few data on improving patient-

physician communication in pain management [27],addressing anticipatory pain with patients before bi-opsies may be of benefit. Our study did not find a directcorrelation between radiologist-patient communicationand biopsy pain, but communication specificallyregarding anticipated biopsy pain may influence itindirectly. Because we found that levels of pain duringbiopsy were significantly lower than anticipated biopsypain levels, simple patient education might reduceanticipatory pain. For example, radiologists could pro-vide information about average pain levels and sensa-tions patients can expect to experience and canencourage patients to use active pain-coping strategiessuch as distraction, relaxation, imagery, or even medi-cations for extreme cases. Although this discussion seemsapplicable immediately before the procedure, it mayhave greater impact if held earlier during the biopsyrecommendation.We obtained pain measurements immediately after

biopsy; we did not measure later retrospective recall ofpain. Although commonly used, retrospective recall maybe inaccurate [28] because the most intense and finalmoments of a pain experience may disproportionatelyinfluence retrospective judgments, introducing bias[29]. Regardless, retrospectively recalled pain may in-fluence later screening behavior, so future studies shouldexamine discrepancies between real-time and retrospec-tive pain recall and their impact on subsequentmammographic screening.Our study had several limitations. First, although

similar techniques were used among patients, theamount of lidocaine administered was not recorded.Second, we did not ask patients about pain history orprocedural factors that might affect pain during biopsy.Consequently, we cannot determine the relative con-tributions of pain caused by the biopsy needle, patientposition during biopsy, procedure length, or preexistingbreast pain. Next, we did not evaluate women’s expe-riences during MRI-guided biopsies or ultrasound-guided biopsies performed with VA devices, nor theuse of hybrid VA multipass devices. Finally, this pilotstudy allowed us to invite only a portion of the clinic’stotal volume, compared with a clinicwide study thatwould have captured all patients. Future studies shouldevaluate these factors.

CONCLUSIONSMost breast biopsy procedures were performed withlittle to no pain, and women who experienced greaterpain levels tended to be those with higher anticipatedpain. Providing patients with education about common

pain levels, sensations they might experience duringbiopsy, and strategies to reduce pain may improve thebiopsy experience.

TAKE-HOME POINTS

� Women undergoing stereotactic breast biopsies usinglarger (smaller gauge) needles reported higher painlevels compared with women undergoing ultrasound-guided procedures using smaller needles.

� Lower patient age, greater breast density, and greaterprebiopsy breast pain were patient-related factorssignificantly associated with biopsy pain.

� Anticipated pain before biopsy and catastrophicthoughts about pain during biopsy were psychologicalfactors strongly associated with pain during the biopsyprocedure.

REFERENCES

1. Whelehan P, Evans A, Wells M, Macgillivray S. The effect ofmammography pain on repeat participation in breast cancer screening: asystematic review. Breast 2013;22:389-94.

2. US Department of Health and Human Services. Medicare program;hospital inpatient value-based purchasing program. Fed Reg 2011:26497-8.

3. Zagouri F, Sergentanis TN, Gounaris A, et al. Pain in different methodsof breast biopsy: emphasis on vacuum-assisted breast biopsy. Breast2008;17:71-5.

4. Zografos GC, Zagouri F, Sergentanis TN, et al. Pain during vacuum-assisted breast biopsy: are there any predictors? Breast 2008;17:592-5.

5. Zografos GC, Zagouri F, Sergentanis TN, Nonni A, Lymperopoulos N,Patsouris E. What parameters affect pain in core biopsy? Eur Radiol2008;18:1144-5.

6. Szynglarewicz B, Matkowski R, Kasprzak P, et al. Pain experienced bypatients during minimal-invasive ultrasound-guided breast biopsy:vacuum-assisted vs core-needle procedure. Eur J Surg Oncol 2011;37:398-403.

7. Satchithananda K, Fernando RAM, Ralleigh G, et al. An audit of pain/discomfort experienced during image-guided breast biopsy procedures.Breast J 2005;11:398-402.

8. Hemmer JM, Kelder JC, van Heesewijk HP. Stereotactic large-coreneedle breast biopsy: analysis of pain and discomfort related to the bi-opsy procedure. Eur Radiol 2008;18:351-4.

9. Denton ERE, Ryan S, Beaconsfield T, Michell MJ. Image-guided breastbiopsy: analysis of pain and discomfort related to technique. Breast1999;8:257-60.

10. Salem C, Sakr R, Chopier J, et al. Pain and complications of directionalvacuum-assisted stereotactic biopsy: comparison of the Mammotome andVacora techniques. Eur J Radiol 2009;72:295-9.

11. Sullivan MJ, Thorn B, Haythornthwaite JA, et al. Theoretical perspec-tives on the relation between catastrophizing and pain. Clin J Pain2001;17:52-64.

12. Al-Obaidi SM, Beattie P, Al-Zoabi B, et al. The relationship of antici-pated pain and fear avoidance beliefs to outcome in patients with chroniclow back pain who are not receiving workers’ compensation. Spine2005;30:1051-7.

13. Edwards RR, Mensing G, Cahalan C, et al. Alteration in pain modula-tion in women with persistent pain after lumpectomy: influence of cat-astrophizing. J Pain Symptom Manage 2013;46:30-42.

14. Rosenstiel AK, Keefe FJ. The use of coping strategies in chronic low-back-pain patients—relationship to patient characteristics and currentadjustment. Pain 1983;17:33-44.

716 Journal of the American College of Radiology/Vol. 11 No. 7 July 2014

15. Keith DJ, Rinchuse DJ, Kennedy M, Zullo T. Effect of text messagefollow-up on patient’s self-reported level of pain and anxiety. AngleOrthod 2010;83:605-10.

16. Miller LS, Shelby RA, Balmadrid MH, et al. Patient anxiety before andimmediately after imaging-guided breast biopsy procedures: impact ofradiologist-patient communication. J Am Coll Radiol 2013;10:423-31.

17. Cleeland CS. Pain assessment: global use of the Brief Pain Inventory.Ann Acad Med Singapore 1991;23:129-38.

18. Spielberger C. Manual for the State-Trait Anxiety Inventory. Palo Alto,California: Mind Garden, 1983.

19. Turner RJ, Wheaton B, Lloyd DA. The epidemiology of social stress. AmSociol Rev 1995;60:104-25.

20. Bieber C, Müller KG, Nicolai J, Hartmann M, Eich W. How does yourdoctor talk with you? Preliminary validation of a brief Patient Self-ReportQuestionnaire on the Quality of Physician-Patient Interaction. J ClinPsychol Med Settings 2010;17:125-36.

21. van Vlymen JM, Sa Rego MM, White PF. Benzodiazepine premed-ication: can it improve outcome in patients undergoing breast biopsyprocedures? Anesthesiology 1999;90:740-7.

22. Bugbee ME, Wellisch DK, Arnott IM, et al. Breast core-needle biopsy:clinical trial of relaxation technique versus medication versus no inter-vention for anxiety reduction. Radiology 2005;234:73-8.

23. Lang EV, Berbaum KS, Faintuch S, et al. Adjunctive self-hypnoticrelaxation for outpatient medical procedures: a prospective randomizedtrial with women undergoing large core breast biopsy. Pain 2006;126:155-64.

24. Kumar M, Kumar J, Saxena I. The role of mental distraction on the painresponse in healthy young Indian adults. J Clin Diagn Res 2012;6:1648-52.

25. Zeidan F, Gordon NS, Merchant J, Goolkasian P. The effects of briefmindfulness meditation training on experimentally induced pain. J Pain2010;11:199-209.

26. Riddle DL, Keefe FJ, Nay WT, McKee D, Attarian DE, Jensen MP.Pain coping skills training for patients with elevated pain catastrophizingwho are scheduled for knee arthroplasty: a quasi-experimental study.Arch Phys Med Rehabil 2011;92:859-65.

27. Pincus T, Holt N, Vogel S, et al. Cognitive and affective reassurance andpatient outcomes in primary care: a systematic review. Pain 2013;154:2407-16.

28. Salé H, Hedman L, Isberg A. Accuracy of patients’ recall of temporo-mandibular joint pain and dysfunction after experiencing whiplashtrauma: a prospective study. J Am Dent Assoc 2010;141:879-86.

29. Schneider S, Stone AA, Schwartz JE, Broderick JE. Peak and end effectsin patients’ daily recall of pain and fatigue: a within-subjects analysis.J Pain 2011;12:228-35.