British Journal of Radiology (2007) 80, 866-871
© 2007 British Institute of Radiology
doi: 10.1259/bjr/11719131
Content analysis of general practitioner-requested lumbar spine X-ray reports
P Thompson, MD, FRCP
1 and
E Carr, RGN, MSC, PhD
2
1 Studland Centre, Poole Hospital NHS Trust, Poole, Dorset BH15 2JB, 2 Institute of Health and Community Studies, Bournemouth University, Bournemouth, Dorset BH1 3LT, UK
Correspondence: Paul Thompson, Consultant Rheumatologist, Poole Hospital NHS Trust, Dorset BH15 2JB. UK. E-mail: paul.thompson{at}poole.nhs.uk
 |
Abstract
|
|---|
X-rays of patients with lower back pain rarely show serious pathology but frequently reveal incidental age-related changes and always expose people to radiation. Patients who have X-rays are more satisfied but report worse pain and disability. Psychological factors such as illness beliefs, catastrophizing and fear avoidance have been shown to be predictors of chronicity/disability. Authorities suggest that the way X-ray information is transmitted and interpreted by patients may influence outcome; therefore, this study was designed to determine the words used by radiologists to describe lumbar spine X-rays. 120 consecutive X-ray reports for patients referred by primary care physicians were anonymized, and a formal summative content analysis undertaken. The coded words were grouped into categories according to their perceived meaning, and the process was refined until there were only three mutually exclusive categories. Half of the sample was aged 60 years or less. Three categories were identified: anatomical, pathological and descriptive. In the pathological category, 33% of words described normal appearances, 47% described age-related changes and 20% described other features. In only 2% of cases were pathological words used to describe conditions as being "normal for age". Overall, 89 (74%) of the 120 reports contained at least one phrase containing words indicating the presence of degenerative changes.
|
Introduction
|
|---|
Non-specific lower back pain is a common problem: over 50% of people will have experienced at least one attack by late middle age [1]. Most episodes are self-limiting but can sometimes develop into long-term disability with enormous cost to the individual and society. Psychological factors such as illness beliefs, catastrophizing and fear-avoidance are important predictors of chronicity and disability [2, 3].
Radiographical studies in urban communities in the 1950s found that "degenerative changes" were common and increased with age but did not correlate with back pain [4]. In a systematic review of 18 studies, Van Tulder et al [5] suggested that degeneration, defined by the presence of narrowing disc space, osteophytes and sclerosis, was associated with back pain. However, because of the high prevalence of degenerative changes and the small associated increased risk of back pain, the actual relationship is weak (see also Symmons et al [6]).
Conversely, a study examining the prevalence of lumbar disc degeneration and lower back pain found that lumbar disc degeneration manifested earlier in a greater number of participants with lower back pain [7]. Therefore, the interpretation of degenerative changes seen on spinal X-rays in the context of a patient with back pain is an uncertain art and remains controversial.
National guidelines advise that routine lumbar spine X-rays are avoided for simple lower back pain because they rarely pick up serious pathology and expose people to radiation [8, 9]; however, patients with back pain want X-rays [10]. Most general practitioners (GPs) order X-rays to reassure their patients [11]. Patients who have spine X-rays are more satisfied with their treatment than patients who do not have X-rays but report more pain, lower overall health status and no difference in disability, and consult their doctor more frequently [12]. Could it be that the way in which X-ray reports are communicated has an effect on the outcome of an episode of back pain? As Waddell states, "Think about the things we tell patients with ordinary backache. Take the example of normal, age-related changes on lumbar spine X-rays: 'you have wear and tear in your spine' or, even worse, 'degenerative disc disease'. To patients, this means serious deterioration; it is irreversible, and will get even worse as they get older. If I'm like this now what will I be like in 10 years? Will I end up in a wheelchair? It is no use saying 'but it is nothing to worry about!'. The damage has been done. We have labelled them with a disease that will make them ill" [13]. We found little in the literature to support or refute the contention that X-ray reports may harm patients, and so wish to explore this further. As a start, we carried out a content analysis of lumbar spine X-ray reports to determine the words used by radiologists to describe radiographical appearances and the frequency of their use.
|
Methods
|
|---|
Having gained Ethics and Research Governance Committee approval and the written informed consent of all relevant radiologists at a district general hospital in the south of England, consecutive reports for GP-requested lumbar spine X-rays were obtained from the senior radiology secretary over a 4-week period. It was anticipated that this would generate sufficient reports to undertake a content analysis. Each report was photocopied and all patient and radiology identifiers removed; the age of the patient was added to the form at this stage.
A formal summative content analysis was undertaken to determine not only the frequency of words [14] but also their contextual meaning [15]. The reports and patients' ages were typed up using Microsoft Word and a word count of each report was undertaken. Each report was then split into individual phrases and these phrases were entered into one column of a spreadsheet (SPSS, version 13; SPSS Inc, Chicago, IL) and numbered. Each word appearing in the phrases was given an exclusive code number that was entered into the next column. Link words such as "and", "the" and "was" were excluded. The coded words were grouped into categories according to their perceived meaning and the process was refined until there were only three exclusive categories labelled in separate columns. For one category, an additional code of "present" or "not present" was added in a separate column, and the code words divided into three subcategories.
Two clinicians (a radiologist and a rheumatologist), blinded to the numerical codes, independently classified all of the coding words into the three categories to assess inter-rater variation.
The spreadsheet enabled us to calculate how frequently different words were used. These results are presented graphically for words occurring five or more times and in a box for words occurring less than five times. Ages in different codes were compared using t-tests, and inter-rater variation was assessed using Cronbach's 
(SPSS, version 13).
|
Results
|
|---|
In total, 120 consecutive GP-requested lumbar spine X-ray reports, dictated by 12 consultant radiologists, were collected and anonymized. The patients' mean (±SD) age was 59 (±18) years with a range of 16–95 years. The age frequency distribution is shown in Figure 1
. Over half of the sample was aged 60 years or less.
The mean (±SD) number of words in each report was 45 (±25); 676 separate phrases were identified with a range from 1–7 for individual reports. 145 code words were identified and grouped into three categories: "anatomical" words (n = 45), "pathological" words (n = 55) and "descriptive" words (n = 45). 86% of reports contained "anatomical" category words, 98% contained "pathological" category words and 14% contained "descriptive" category words. Inter-rater agreement was good (Chronbach's 0.85).
Figure 2 shows the frequency of each code word for the 495 words in the "anatomical" category. Words relating to the intervertebral disc were most commonly used.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 2. Frequency of words in the"anatomical" category in 120 reports of GP-requested lumbar spine radiographs occurring either five times or more (shown in the graph) or less than five times (in the box).
|
|
Figure 3 shows the frequency of each code word for the 655 words in the "pathological" category (present = 594; not present = 59). The code words have been further classified into three subcategories: (a) "normal appearances" (n = 181); (b) "degenerative changes" (n = 269 present, n = 14 absent); and (c) "other features" (n = 146 present, n = 45 absent). Overall, 74% (n = 89) of the 120 reports included at least one phrase containing code words in the "degenerative changes" subcategory.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 3. Frequency of words in the"pathological" category either "present" or "not present" in 120 GP-requested lumbar spine X-ray reports occurring either five times or more (shown in the graph) or less than five times (in the box).
|
|
The words in the "normal appearances" subcategory related to "anatomical" category words in most cases, e.g. "normal vertebral height" and "disc spaces maintained". In only 2% of cases were changes described as being "normal for age". Only a small number of words indicated the possibility of serious pathology, e.g. "fracture", "lesion" and "anomaly", although this was usually mentioned for its absence, as in "no anomaly seen". The frequency of words in the "normal appearances" subcategory fell from 68% in the 21–30 year age group to 17% in the over 80s, whereas the frequency of words in the "degenerative changes" subcategory rose from 4% in the 21–30 year age group to 49% in the over 60s.
Figure 4 shows the frequency of each code word for the 205 words in the "descriptive" category. These words mainly related to the degree of "pathological" category changes, e.g. "mild degenerative arthritis". In most cases, severity was described with words meaning mild or moderate.
View larger version (21K):
[in this window]
[in a new window]
|
Figure 4. Frequency of words in the"descriptive" category in 120 reports of GP-requested lumbar spine radiographs occurring either five times or more (shown in the graph) or less than five times (in the box).
|
|
|
Discussion
|
|---|
This was a study of normal practice during 1 month in a busy district general hospital. We were surprised at the number of GP requests for X-rays for patients with lower back pain (>30 per week) from a catchment population of about 200 000, although the numbers were similar to those reported at the neighbouring Royal Bournemouth Hospital. We were unable to assess how many of the referrals fulfilled national guidelines with respect to red flags [8, 16] but it would appear that there is considerable pressure on GPs to order X-rays. Indeed, a recent Norwegian study assessing the myths and perceptions of back pain in a sample of the general population (n = 1014) confirmed that the use of X-rays to identify the cause of pain and the necessity to have a spine X-ray are deeply entrenched beliefs [17].
Post-mortem studies have demonstrated morphological changes in lumbar spine segments; the discs appear hardened, cracked and narrowed; bony spurs (osteophytes) form around the disc margins; and bone is laid down under endplates (sclerosis). Cartilage of the facet joint fragments and osteophytes form, giving a hypertrophied appearance. These changes are more common in men and are found more frequently in the lower lumbar segments [18]. They can be detected on lumbar spine X-rays and MRI scans, and are frequently summarized as "degenerative changes". The X-ray appearances correlate with the changes seen at autopsy [19] and their prevalence increases with age — rising, in one study, from 20% in under 35-year-olds to 71% in 65- to 74-year-olds [20]. Although the frequency of degenerative changes was lower in younger people, this age group's interpretation of the words "degenerative" and "spondylosis" may have greater impact compared with that of people in their 70s. In our study, a significant number of patients were young: 20 patients were under 40 years of age and just over half of the sample was aged 60 years or less, although this age range is similar to other studies [20]. Perhaps "uncommon for age" would be a more useful description, as "age-related" changes are by definition uncommon (and therefore potentially abnormal) in young individuals.
It is already known that the influence of psychological factors appears increasingly important in the transition from acute to chronic lower back pain [21, 22]. Indeed, the potential role of "catastrophizing" has received considerable interest. The contemporary view is that catastrophizing involves rumination, magnification and helplessness [2]. It has emerged as one of the most robust predictors of the pain experience [23] and has been observed in chronic pain patients [24]. At this point, we can only speculate on the role that catastrophizing may play for patients receiving their X-ray report findings. We could find no studies that explore how the findings from X-ray reports are explained to the patient and this warrants further exploration.
To our knowledge, our study is the first formal content analysis of lumbar spine X-ray reports, although other studies have shown a similarly high prevalence of degenerative changes that correlate positively with age, and normal findings that correlate negatively with age [19]. There were, however, several limitations including ethical review, sample size, validity and generalizability.
Due to Ethics Committee constraints, we were requested to remove all patient information (except age), as well as the radiologists' identities. We were also unable to collect the patients' demographic or clinical data. Constraints and variation in ethical review and the negative impact this may have on research has been the subject of much debate [25]. The impact of these decisions may have resulted in unequal contributions to the sample of reports by individual radiologists, although we know that all 12 were reporting during the study period. Although we cannot rule out any sex bias or patient selection on clinical grounds, e.g. the use of red flags, Van den Bosch et al [20] found a similar frequency of degenerative changes in both men and women.
Determining an adequate sample size for content analysis was achieved by reviewing similar studies. We found wide variation in sample size, from 81 e-mail communications between a primary care provider and patients [26] to 921 newspaper articles to understand how print coverage may affect primary and secondary skin cancer prevention in the USA [27]. The traditional inference that an increased sample size improves validity and reliability is not directly transferable to the analysis of qualitative data. Our results for frequency of findings and relationship to age are similar to other studies, which supports the adequacy of the sample.
In qualitative research, the term validity is replaced by credibility, which deals with the focus of the research and how confidently data and data processes address the intended focus of the study. Concerns regarding the generalizability of the results or findings are replaced with transferability. Transferability refers to the extent to which the results can be generalized or transferred to other contexts or settings, and is primarily the responsibility of the individual doing the generalizing. We believe this study demonstrates this in that the findings echo those of our peers [13] and the published literature [20]. Qualitative researchers can enhance transferability by thoroughly describing the research context and the assumptions that were central to the research study, and by ensuring that they are clearly articulated in the paper [28].
|
Conclusions
|
|---|
This study showed that three-quarters of lumbar spine X-ray reports included at least one phrase containing words indicating the presence of "degenerative changes". It was rare for the radiologists to comment that the changes were "normal for age". Although we cannot make assumptions about how emotive the use of these words might be in the context of an episode of back pain, the study indicates that there is the possibility for patients to misinterpret their report as described by Waddell [13]. Recognizing this, Roland and Van Tulder [29] proposed that radiologists should use epidemiological information to help interpret their findings, with a view to reducing the potential for misinterpretation of their reports. None of the reports we analysed contained this type of information. Further studies are needed to determine how GPs interpret the results, how they explain these results to patients, how patients then interpret them, what actions the patients take and, most importantly, the effect that these have on the persistence and severity of pain and disability.
|
Acknowledgments
|
|---|
We thank Tracey Wellstead for preparing the X-ray reports and Professor Alan Breen for his helpful and constructive comments on the manuscript.
Received for publication June 1, 2006.
Revision received January 24, 2007.
Accepted for publication February 5, 2007.
|
References
|
|---|
- Papageorgiou AC, Croft PR, Ferry S, Jayson MIV, Silman AJ. Estimating the prevalence of low back pain in the general population. Evidence from the South Manchester back pain survey. Spine 1995;20:1889–94.[Medline]
- Sullivan MJL, Bishop SR, Pivik J. The Pain Catastrophizing Scale: development and validation. Psychol Assess 1995;7:524–32.[CrossRef]
- Linton SJ. A review of psychological risk factors in back and neck pain. Spine 2000;25:1148–56.[CrossRef][Medline]
- Kellgren JH, Lawrence JS. Osteo-arthrosis and disk degeneration in an urban population. Ann Rheum Dis 1958;17:388–97.[Free Full Text]
- Van Tulder MW, Assendlft WJ, Koes BW, Boulter LM. Spinal radiographic findings and nonspecific low back pain. A systematic review of observational studies. Spine 1997;22:427–34.[CrossRef][Medline]
- Symmons DP, van Hemert AM, Vandenbroucke JP, Valkenburg HA. A longitudinal study of back pain and radiological changes in the lumbar spines of middle aged women II. Radiographic findings. Ann Rheum Dis 1991;50:162–6.[Abstract/Free Full Text]
- Paajanen H, Erkintalo M, Parkkola R, Salminen J, Kormano M. Age-dependent correlation of low-back pain and lumbar disc degeneration. Arch Orthop Trauma Surg 1997;116:106–7.[CrossRef][Medline]
- Koes BW, van Tulder MW, Ostelo R, Burton AK, Waddell G. Clinical guidelines for the management of low back pain in primary care: an international comparison. Spine 2001;26:2504–14.[CrossRef][Medline]
- ACC. New Zealand Acute Low Back Pain Guide [online]. Wellington, New Zealand: ACC; 2004. Available from http://www.nzgg.org.nz/guidelines/0072/acc1038_col.pdf [Accessed 4 January 2006]
- Klaber Moffett JA, Newbronner E, Waddell G, Croucher K, Spear S. Public perceptions about low back pain and its management: a gap between expectations and reality? Health Expectations 2000;3:161–8.[CrossRef][Medline]
- Kaplan DM, Knapp M, Romm FJ, Velez R. Low back pain and X-ray films of the lumbar spine: a prospective study in primary care. South Med J 1986;79:811–4.[Medline]
- Kendrick D, Fielding K, Bentley E, Miller P, Kerslake R, Pringle M. The role of radiology in primary care patients with low back pain of at least 6 weeks' duration: a randomised (unblinded) controlled trial. Health Technol Assess 2001;5:1–69.[Medline]
- Waddell G. The Back Pain Revolution, 2nd edition. London: Churchill Livingstone; 2004
- Manning PK, Cullum-Swan B. Narrative, content and semiotic analysis. In: Denzin NK, Lincoln YS, editors. Handbook of Qualitative Research. Thousand Oaks, CA: Sage Publications; 1994: 463–7
- Hsieh HF, Shannon S. Three approaches to qualitative content analysis. Qual Health Res 2005;15:1277–88.[Abstract/Free Full Text]
- Van Tulder M, Becker A, Bekkering T, Breen A, Gil del Real MT, Hutchinson A et al, on behalf of the COST B13 Working Group on Guidelines for the Management of Acute Low Back Pain in Primary Care. European guidelines for the management of acute nonspecific low back pain in primary care [online]; 2004. Available from http://www.backpaineurope.org/web/files/WG1_Guidelines.pdf [Accessed 4 January 2006]
- Ihlebaek C, Eriksen HR. Myths and perceptions of back pain in the Norwegian population, before and after the introduction of guidelines for acute back pain. Scand J Pub Health 2005;33:401–6.[CrossRef]
- Miller JA, Schmatz C, Schultz AB. Lumbar disc degeneration: correlation with age, sex, and spine level in 600 autopsy specimens. Spine 1988;13:173–8.[Medline]
- Benneker LM, Heini PF, Anderson SE, Alini M, Ito K. Correlation of radiographic and MRI parameters to morphological and biochemical assessment of intervertebral disc degeneration. Eur Spine J 2005;14:27–35.[CrossRef][Medline]
- Van den Bosch MAAJ, Hollingworth W, Kinmouth AL, Dixon AK. Evidence against the use of lumbar spine radiography for low back pain. Clin Radiol 2004;59:69–76.[CrossRef][Medline]
- Pincus T, Burton AK, Vogel S, Hilderbrandt VH. A systematic review of psychological factors as predictors of chronicity/disability in population cohorts of low back pain. Spine 2002;27:E109–20.[CrossRef][Medline]
- Linton SJ. Do psychological factors increase the risk for back pain in the general population in both a cross-sectional and prospective analysis? Eur J Pain 2004;9:355–61.[CrossRef]
- Geisser ME, Robinson ME, Riley JL. Pain beliefs, coping, and adjustment to chronic pain. Pain Forum 1999;8:161–8.
- Spinhoven P, Ter Kuile MM, Linssen ACG, Gazerdam B. Pain coping strategies in a Dutch population of chronic low back pain patients. Pain 1989;37:77–83.[CrossRef][Medline]
- Shaw S, Boynton PM, Greenhalgh T. Research governance: where did it come from, what does it mean? J Roy Soc Med 2005;98:496–502.[Abstract/Free Full Text]
- Anand SG, Feldman MJ, Geller DS, Bisbee A, Bauchner H. A content analysis of e-mail communication between primary care providers and parents. Pediatrics 2005;115:1283–8.[Abstract/Free Full Text]
- Stryker JE, Solky BA, Emmons KM. A content analysis of news coverage of skin cancer prevention and detection, 1979 to 2003. Arch Dermatol 2005;141:491–6.[Abstract/Free Full Text]
- Greenhalgh T, Taylor R. How to read a paper: papers that go beyond numbers (qualitative research). BMJ 1997;7110:740–3.
- Roland M, van Tulder M. Should radiologists change the way they report plain radiography of the spine? Lancet 1998;352:229–30.[CrossRef][Medline]
Top
Abstract
Introduction
