There are several ways in which we might assess the importance of pain in mammography as a problem. We could ask what proportion of women undergoing mammography experience any pain, or what proportion of women experience pain above a specified level. The literature in this area provides vastly variable findings, largely because of methodological limitations. Literature reviews have found that the proportion of women experiencing pain during their mammograms ranges from 1 to 92 % [7, 8], or from 6 to 76 % [9], so it is clearly difficult to use prevalence rates as a measure of the importance of the problem.

Perhaps a more appropriate measure of the importance of pain in mammography is whether, in the breast screening context, it affects behaviour; i.e. does it deter women from returning for future mammographic screening. Findings in the literature have varied on this question but a recent systematic literature review [10] has established that between 25 and 46 % of breast screening non-re-attenders give pain as a reason for not returning. This review also showed that when pain is measured at an index mammogram and compared with subsequent re-attendance rates, the risk of non-re-attendance is about a third higher in women who report pain than in those who do not (risk ratio: 1.34 [95 % CI: 0.94–1.91]). In the context of surveillance mammography for women previously treated for breast cancer, a study in 2012 [11] did not find an association between pain at mammography and annual mammography adherence. However, it was demonstrated that anxiety about the mammogram, and pain catastrophizing (e.g. “I became afraid that the pain would get worse”) were associated with non-adherence.

This is another question which many authors have attempted to address. Numerous effects have been implicated as risk factors for pain in mammography but the evidence is inconclusive for many of them. This is partly because of weaknesses in the methodologies used in some studies, including the use of non-validated pain measurement instruments. In addition, it is difficult to separate the many potential co-variables and the complex interactions between them which are likely to exist.

An informal literature review published in 2007 [8] grouped the risk factors for mammography pain into biological, psychological and staff-related. Biological factors that have been linked with greater pain include breast tenderness; psychological factors include expectations of pain, and staff-related factors include clients’ perceptions of staff attitude.

There is, perhaps, a fourth important category – technique-related. There is very little empirical evidence that the specific equipment model has an influence on mammographic pain, although many practitioners will suggest that it does. However, another obvious place to look for associations between technique and pain is the compression force exerted on the breast. This was investigated by Sullivan et al in 1991[12], whose findings suggested that pain was related to compression force. As it is recognised that applied force varies by practitioner [13], a strength of this study was that a single practitioner x-rayed all the participants. However, it is not stated that the cohort in the Sullivan study was asymptomatic (the age range would suggest not), nor was any differential presence of symptoms taken into account. Furthermore, the scale used to assess pain was non-standard and no evidence of validity or reliability was provided. A study by Poulos and Rickard later found no difference in discomfort between two cranio-caudal views when one was deliberately compressed less firmly than the other by the practitioner [14]. Clearly, a lack of robust empirical evidence for a relationship between compression force and pain does not necessarily mean that no relationship exists. The advent of digital mammography and tools for automated extraction of technique data may facilitate larger-scale and more definitive studies in this area.

Compression, and other aspects of technique, are discussed further in the next paragraph.

Clearly, we should target those potentially modifiable factors which contribute most to the problem but, as described above, a surprising lack of clarity persists regarding what those factors are. A Cochrane systematic review, last updated in 2008 [15], found a shortage of effective interventions to reduce mammography pain. Interventions showing most promise in randomised controlled trials were: giving women sufficient information about the procedure prior to the mammogram; increasing women’s control over the level of compression applied; and use of cushioning on the mammography machine. However, the latter two interventions both carried the risk of detriment to image quality, and the cushions, at least, involve additional cost.

An obvious potential pain reduction intervention is medication, and a number of studies have been conducted in this area. A well-conducted multi-arm, randomised, placebo controlled trial [16], which was published slightly too late to be included in the 2008 Cochrane review, evaluated the effects of lidocaine gel application and oral premedication with ibuprofen or paracetamol on mammography discomfort and satisfaction. The authors found a statistically significant but very small, and therefore probably clinically insignificant, difference in reported discomfort for lidocaine gel compared with placebo or no gel.

While researchers continue the quest for feasible, effective and cost-effective interventions to reduce pain in mammography, there are measures that all practitioners can take in their daily work which may reduce pain or discomfort and/or increase client satisfaction, without risk of causing harm or incurring additional costs. Provision of sufficient information and explanation should of course always be part of standard practice. In addition, it was demonstrated in one study that the risk of women reporting pain from mammography was reduced if they perceived that the radiographer told them that they could say “stop” if they became too uncomfortable, and if they perceived that they had had a conversation with the radiographer [17]. The method of assessing pain is not clearly described in this publication and there is no evidence of validity testing having been performed on the questionnaires. However, the finding of reduced pain risk if women are verbally offered some control over the level of compression is consistent with the randomised controlled trial evidence that pain can be reduced by giving more control to the women [18].