Most people will suffer from low back pain (LBP) at least once in their life, and many of them will experience more than one period of LBP. 46 The pain may subside and disappear for a while and then recur or reappear a few months or years later. The pain may also linger for some time and flare up periodically. If these flare-ups are bothersome, this may prompt the patient to seek medical care or to have time off work. Low back pain can therefore be characterized as an episodic disease. Unfortunately, available data on the natural history and clinical course of LBP are often incomplete and confusing. Nachemson and Bigos 28 proposed definitions for acute and chronic LBP, which Von Korff 45 refined by clarifying ambiguities and enlarging the taxonomy. This was done by defining transient back pain, recurrent back pain, chronic back pain, acute back pain, first onset, and flare-up. These definitions are important for both clinicians and researchers interested in the onset, prognosis, and clinical course of LBP. In this set of definitions, however, an important one for researchers is missing; namely, the definition of an episode.
There are several situations for which a clear definition of an episode is needed. First, it is common in cohort studies to focus on the development of new episodes of back pain. In theory, cohort studies examining risk factors for the occurrence of a specific disease require patients who are free of the disease at inclusion. Studying the risk factors for LBP, a population that has never had LBP before is not only a small population but is also likely to be a very young one. Because LBP is an episodic disease, it is more appropriate to study patients who have been free of LBP for a specific period and who can thus be considered at risk for a new episode of LBP. Second, in randomized clinical trials, one sometimes wishes to study acute or subacute patients. The important question in the identification of these patients is when the current episode actually started, and whether the preceding period was indeed free of LBP. Third, in long-term follow-up in randomized clinical trials and cohort studies, we need to monitor symptoms over time and identify and distinguish resolution, persistence, and recurrence of symptoms, as well as the occurrence of new episodes. In this situation also, a proper definition of an episode is essential.
So several fundamental aspects of studying LBP rely on the identification of episodes. Or as Buckle 3 puts it: “The frequency of attacks and the number of episodes of back pain can create problems for the researchers” (p. 320). To improve the understanding of the etiology, clinical course, and prognosis of back pain, standardized operational definitions of episodes need to be applied in longitudinal research. For that reason, we have attempted to define episodes of LBP in a way that is tenable and applicable in research and in clinical practice.
In this article, we first present our findings from the literature review on explicitly and implicitly used definitions in papers on episodes of LBP, episodes of care for LBP, and episodes of work absence because of LBP. Then we consider the proposal of uniform definitions of episodes and discuss applications and implications for future standardized use.
The literature search aimed to identify papers in which episodes of LBP were defined and/or in which specific definitions of episodes of LBP were used. First we searched in Medline (1966–2000) using a combination of the MESH terms “back pain,” “backache,” or “back ache” with “episod*,” “period*,” “recurr*,” “relapse,” “intermitt*,” “inciden*,” “course,” “interval,” “initial*,” or “consult*” with “defin*,” “descript*,” “describ*,” “classif*,” “concept*,” “categoris*,” or “categoriz*.” This led to 890 papers. The titles and abstracts were examined by two reviewers independently (H.C.W.d.V. and M.W.H.) on the probability that the study contained a definition of episodes of LBP. In case of discrepancy or doubt, the study was included. This led to 37 potentially relevant papers. The second search strategy (Medline 1966–2000) consisted of the terms “back pain” and “episodes” in the titles or abstracts and identified 349 papers. Using the same procedure, 44 additional papers were included. Furthermore, textbooks 2,11,29,47 on LBP were screened, and the references in all retrieved papers were checked for additional papers in which explicit definitions of episodes of low back could be found. This did not yield additional papers.
The following information was extracted from the papers: the definition of an episode if presented or if derivable, study setting, type of back pain, minimum duration of the LBP-free period before entry into a study, minimum period between two distinct episodes, duration of the episodes, outcome measure used, and intensity of LBP. Data were extracted independently from each of the 81 papers by two of four authors (H.C.W.d.V., M.W.H., K.M.D., D.P.P.). Afterwards, the results were compared, and consensus was reached by discussion. All reviewers and data extractors are experienced researchers in the field of LBP. The papers were grouped into three categories. The first category concerned papers with data about LBP, usually obtained by interview or questionnaire from the patients. The second category consisted of papers presenting data about care for LBP, obtained either from health care registries or by inquiry of the patient. The third category dealt with papers with data on work absenteeism due to LBP, derived from registries or reported by the patients or occupational physicians. Some papers fit into more than one category.
The data from the literature formed the starting point for the group discussions. In this way, a large number of international researchers indirectly gave input to the discussions. In the group meetings, in which all authors participated, the different definitions were proposed, and it was discussed how tenable and applicable they were.
Literature Data on Episodes
Of the 81 papers examined, only 31 contained a (mostly implicit) definition of episodes of LBP (references of the other papers can be obtained on request from the first author). The data from these 31 papers are presented in Table 1 for 16 studies on LBP, 5–10,14,22–24,31–33,35,44,48 in Table 2 for 9 studies on care for LBP, 12,17,25,27,36–39,42 and in Table 3 for 6 studies on work absenteeism due to LBP. 1,16,19–21,30 In these tables, we distinguish between definitions focusing on the minimum duration of the LBP-free period before the start of a new episode (e.g., for inclusion of a cohort study or a randomized clinical trial) and definitions focusing on the minimum duration of the LBP-free period between two episodes in longitudinal studies on the course of LBP. The tables also show the minimum duration of the pain-free period, how these data were assessed, and further information on the design and the aim of the study.
There was a large variation in the minimum duration of LBP-free periods used to define episodes. The choice for one or another (implicit) definition of an episode was never supported by arguments or scientific considerations, making them all arbitrary. In some categories, the number of studies was too small to conclude what minimum duration of LBP-free period is typically used. Studies in occupational epidemiology often made use of registries. In these, the characteristics and details of the different registries determined which data were available to define episodes.
Considerations on Episodes From the Literature
Given the heterogeneity of the definitions identified from the literature review, it was impossible to recommend a uniform definition from the literature alone. Nevertheless, examination of the literature was helpful because some papers provided arguments for the choices they had made, even if these were typically described as arbitrary. 25,37–39 If we found arguments in the papers, it was usually for not distinguishing different episodes, as illustrated by the following citations. Abenhaim et al1 stated that “there is no objective way to determine if a given episode of back symptoms is independent or not from a previous injury. In the absence of such evidence, it would be equally appropriate to consider every episode of back pain as a recurrence (relapsing symptoms for a previous injury) or as a new episode (independent from any previous medical history)” (p. 831). Smedley et al41 remarked, “...even with more complete information, there would have been difficulty in defining exactly when an episode of back pain has ended and subsequent symptoms represented a new episode rather than a continuation of the earlier illnesses” (p. 2425). And for the occupational setting, Infante-Rivard and Lortie 20 stated, “In our study we are not sure whether all relapses were new episodes, although average time to return to work from onset of treatment was very long (126 days), the shape of the cumulative survival curve suggested that some relapses or interruptions may have been due to untimely returns...” (p. 333). Krause et al21 remarked, “Researchers, therefore, have to choose from several options of creating outcome measures which are determined by conceptual considerations, kind and availability of administrative data, and data management resources” (p. 605).
Garcy et al16 suggested that an injury in the same spinal anatomic area could be considered as recurrent back pain and in another spinal anatomic area as a new episode.
The results presented in Tables 1 to 3 formed the starting point for the group discussions. Because the literature data were sparse and showed a large variation, arbitrary choices had to be made. Six discussion sessions among the authors (at least four participating each time) were necessary to come to resulting definitions, which are presented in Figure 1. Topics during these meetings were as follows: considerations of arguments found in literature on episodes (described above); the justification of the definitions, in terms of duration of the pain-free period, the correspondence (or lack of it) between the three definitions; the focus on pain or disability; and considerations of feasibility and applicability in research and in clinical practice.
Justification of the Definitions
We started with a definition of an episode of LBP as a period of LBP preceded and followed by 1 month without LBP. The period of 1 month is regularly used in research and is also a realistic option taking the limited ability of pain recall by the patients into account.
When focusing on episodes of care, we considered a 3-month gap between two episodes reasonable, although we acknowledge that there are patients with chronic LBP who do not seek medical care every 3 months. It is known that most patients do not seek medical care immediately when they experience LBP. 8 Moreover, the care provider typically does not know when an episode of LBP ends, as the patient will usually not visit him when the complaints have disappeared or may cease to visit despite having pain. Therefore, the duration of the LBP-free period for “pain-based” definitions of an episode ought to be shorter than the corresponding “care-based” episode. Choosing a period longer than 3 months between two different episodes would still not solve the problem of nonconsulting patients with persistent LBP. Furthermore, the existence of waiting lists for advanced diagnostic procedures or treatments of more than 3 months may lead to misclassification if the data collection is register based.
It is important to note that “care-based” episodes are not intended to replicate “pain-based” episodes but exclusively focus on episodes of LBP for which medical care was sought. Many determinants of seeking care can be considered, pain being only one of these. The “care-based” definition is applicable to all forms of care for LBP that are registered.
For an episode of work absence due to LBP, we had difficulties in defining a reasonable minimum absence-free period. Work absence data are very often extracted from registries that contain no or minimal information about the pain, the disability, or the clinical characteristics of the patient. We conducted a small survey of social security systems in a few countries and found that the way in which work absence registries are set up differs substantially between countries and depends on the workers’ compensation system in place. From the literature review, it appeared that studies on work absence using registry data 1,21,30 required only 1 or a few days of return to work to separate two episodes of work absence due to LBP, although 1 or a few days back at work is possibly a failed “return to work” trial. Patients with back pain go on sick leave and remain at home for other reasons than (only) their back pain. On the contrary, many patients with LBP will not take time off work. 15 Hence, even periods of work absence of 3 months apart may be due to the same underlying episode of LBP. Obviously, work absence due to LBP is not a good proxy for an episode of LBP, as work absence is determined by a large number of factors other than pain. If we had chosen 1, 2, or 4 weeks of return to work in between two episodes, it might be suggested that the absence-based episodes reflect episodes of LBP. With the choice of a minimum of 1 day of work resumption to demarcate two distinct episodes of work absence due to LBP, it is clear that an episode of work absence due to LBP does not replicate or stand as a proxy for an episode of LBP.
The proposed definition of episode of work absence implies that the episode of work absence due to LBP continues as long as a patient remains disabled for work. If a patient with LBP returns to work but ends up in a less demanding job or with adaptations at the workplace, the work absence due to back pain ends at the moment the patient starts his new ultimate job.
Focus on Episodes of Back Pain or Back Disability
We had also extracted data from the papers about the outcome measures on which an episode definition was based, such as pain or disability, and its intensity. We rejected the inclusion of intensity into the definition because we did not identify any paper that made use of the intensity of pain to define episodes of LBP. In the case of care utilization and work absenteeism, it is assumed that the complaints are severe enough to prompt medical consultation or sick leave, respectively.
We extensively discussed whether we should base the definitions on LBP or on disability due to back pain. Both are relevant phenomena for the patient. We thought that asking about “disabling back pain,” or pain that limits daily activities, increases the complexity for the patient. It is less ambiguous for the patient to state whether they have had pain during a specific period of time. Therefore, we decided to use only the term “low back pain.”
Considerations of Feasibility and Applicability
We introduced some pragmatic considerations to the discussion. We rejected the idea by Garcy et al16 of including anatomic site into the definition because patients certainly might have new episodes in the same anatomic location and because it is difficult (especially for patients) to determine whether the same anatomic region is involved.
With respect to the choice of minimum duration of the LBP-free period before and after an episode, we considered the ability of patients to recall previous pain. The choice of “1 day back in original work” to distinguish two episodes of work absence, for example, was partly driven by practical concerns. In most studies, it is known whether a worker has resumed his original job but not always the exact period of resumption. Finally, we compared our definitions with what, to our knowledge, occurs in clinical practice, to determine the potential external validity of future studies using the proposed definitions.
The need for uniform definitions for episodes of LBP was underlined in three ways. First, there was a lack of explicit definitions of LBP episodes in the literature. In some papers using data from registries, episodes or recurrences were defined ad hoc for the purpose of the study at issue. 1,21,37–39 Second, we found a substantial number of studies that reported on episodes without defining them. 4,14,18,26,40,43 Some of these asked patients to report the number of episodes they had had in a previous specified period, without presenting any definition. This approach essentially uses the individual patients’ ideas of what constitutes an “episode.” However, patients may have very different perceptions of this, which makes the answers to this question unsatisfactory for clinical and epidemiologic studies. Third, some papers stress the need for methods to improve the description of the course of LBP. Von Korff 45 stated, “Improved information on the natural history of back pain is needed to enable doctors and their patients to understand the likely course of back pain” (p. 2045S). Moreover, different etiologies and causes add to the complexity of definition of recurrent back pain and distinct episodes. In addition, Smith and Stano 42 remarked, “Further research needs to more carefully address problems of identifying and separating episodes because the methods can substantially affect the results” (p. 10).
If episodes were used in the literature, they were applied in many different ways. None of the definitions was supported by scientific arguments, implying that they were all arbitrary or pragmatic choices.
Our literature search yielded a large number of articles. However, it is possible that some papers providing explicit definitions of LBP episodes were missed. This might be the case, especially for definitions given little emphasis in the article at issue. However, we did not find further references in any of the retrieved papers to other published explicit definitions of episodes. We inevitably have missed a large number of studies using implicit definitions of episodes of LBP. In most randomized clinical trials and cohort studies on LBP, it is likely that such implicit definitions have been used. If the authors have put emphasis on episodes, we probably retrieved the paper, while missing an unknown number of others. We conclude that the literature has given us a reasonable impression of what types of definitions of episodes have been used in studies on LBP.
The definitions that we propose for an episode of LBP are most applicable in patients who do indeed have clear periods of LBP, alternating with LBP-free periods. For persons with chronic LBP who often do not experience LBP-free periods at all, episodes according to our definition cannot be identified. In those situations, we suggest use of the term “flare-up,” according to the definition proposed by Von Korff. 45 He defined a flare-up of back pain as a phase of pain superimposed on a recurrent or chronic course. A flare-up refers to a period (usually 1 week or less) when back pain is markedly more severe than is usual for the patient at issue.
Furthermore, Von Korff 45 defined chronic back pain as back pain present on at least half of the days in a 12-month period in a single or in multiple episodes; recurrent back pain as back pain present on less than half of the days in a 12-month period, occurring in multiple episodes over the year; transient back pain as an episode in which back pain is present on no more than 90 consecutive days and which does not recur over a 12-month observation period; and acute back pain as pain that is not recurrent or chronic (as defined above) and whose onset is recent and sudden. The LBP-free periods inherently included in these definitions are quite long. For example, in transient back pain, the definition requires the patient to be free of pain during 12 months. We think that considerably shorter disease-free periods should be used for distinguishing different episodes. Thus, multiple episodes, according to our proposed definitions, fit in a course of recurrent and chronic LBP, as defined by Von Korff. 45
As mentioned earlier, we opted to ask about LBP instead of LBP-related disability. This does not preclude an additional assessment of disability. There might be situations in which there are strong arguments for focusing on back pain-related disability. Clear instructions for the patient on how to interpret such a question are needed. Also when inquiring about LBP, it is important to ensure that the obtained information is both reliable and valid. Prespecification of the region of interest, e.g., by the use of mannequins, is advocated rather than putting the responsibility of identifying the back region on the patients. 34 The longer the period over which information is asked retrospectively, the less likely the responses will be valid, given inaccuracy of recall. Suppose, for example, that patients with LBP for between 2 weeks and 3 months duration are included in a study. These patients have LBP for a relatively short period only. We assume it to be quite feasible for such patients to answer the question: “How long is it since you had a whole month without any back pain?”
Studies on LBP are difficult to compare because of different definitions of episodes. This paper has proposed uniform definitions for episodes and recommends their use in future research on LBP. This will make studies on LBP more comparable in the future. However, this is no substitute for the task of carefully monitoring the course of back pain, together with the disability it causes, the medical consultations to which it leads, and the accompanying periods of work absence.
Until now, the episodic nature of LBP has often been ignored in epidemiologic research, not only by the lack of clear definitions for episodes of LBP, episodes of care, or episodes of work absence, but also by the design of the studies. In cohort studies, a disease-free population is usually chosen to identify risk factors for LBP. Because of its episodic nature, patients with LBP might also be included in the population at risk because they may continue to experience LBP over time, become pain free, or develop new episodes of LBP. Similarly, cohort studies on prognostic factors may include patients who have experienced an episode of LBP previously but at the start of the cohort study are free of LBP. They are again at risk of experiencing a next episode. Investigators should conceptualize LBP as a dynamic process, monitoring the clinical course of the disease over time rather than only performing measurements of pain and/or disability at specific moments. The analysis of these studies should also take this episodic nature into account. This means that simple survival analysis, to examine time until the next episode, or logistic regression analysis to evaluate the presence of low back at a certain moment, considers only part of the LBP problem. To study the clinical course with its episodes, and eventually flare-ups, more advanced designs and statistical methods are required. 13
The proposed, uniform definitions are arbitrary but well-considered definitions of episodes. We do not suggest that all researchers should use these definitions without regard to scientific and pragmatic considerations. But if they have no good arguments for another choice, they may use the definitions proposed here. We recommend their use in research on the course of LBP and simultaneous evaluation of their applicability. The proposed definitions draw explicit attention to the episodic nature of the course of LBP and hopefully bring an end to the heterogeneity in arbitrary definitions of episodes.
- Few explicit definitions of episodes of low back pain were found in the literature.
- This article proposes three definitions: for an episode of low back pain, for an episode of care for low back pain, and for an episode of work absence resulting from low back pain.
- These definitions will hopefully lead to a more uniform and sensible use of the concept of episodes in future research.
The authors thank Rienk Prins, AS/tri, The Netherlands, for inquiring in different countries about definitions of work absence.
1. Abenhaim L, Suissa S, Rossignol M. Risk of recurrence of occupational back pain over three year follow up. Br J Ind Med 1988; 45: 829–33.
2. Anderson GBJ. The Epidemiology of Spinal Disorders. 2nd ed. New York, NY: Raven Press; 1997.
3. Buckle P. Epidemiological aspects of back pain within the nursing profession. Int J Nurs Stud 1987; 24: 319–24.
4. Burton AK, Tillotson KM, Troup JD. Prediction of low-back trouble frequency in a working population. Spine 1989; 14: 939–46.
5. Carey TS, Garrett JM, Jackman A, et al. Recurrence and care
seeking after acute back pain: results of a long-term follow-up study. North Carolina Back Pain Project. Med Care
1999; 37: 157–64.
6. Carey TS, Garrett JM, Jackman AM. Beyond the good prognosis: examination of an inception cohort of patients with chronic low back pain
. Spine 2000; 25: 115–20.
7. Coste J, Delecoeuillerie G Cohen de Lara A, et al. Clinical course
and prognostic factors in acute low back pain
: an inception cohort study in primary care
practice. Br Med J 1994; 308: 577–80.
8. Croft PR, Macfarlane GJ, Papageorgiou AC, et al. Outcome of low back pain
in general practice: a prospective study. Br Med J 1998; 316: 1356–9.
9. Croft PR, Papageorgiou AC, Ferry S, et al. Psychologic distress and low back pain
: evidence from a prospective study in the general population. Spine 1995; 20: 2731–7.
10. Croft PR, Papageorgiou AC, Thomas E, et al. Short-term physical risk factors for new episodes
of low back pain
: prospective evidence from the South Manchester Back Pain Study. Spine 1999; 24: 1556–61.
11. Crombie IK, Croft PR, Linton SJ, et al. eds. Epidemiology of Pain. Seattle: IASP Press, 1999.
12. Dettori JR, Bullock SH, Sutlive TG, et al. The effects of spinal flexion and extension exercises and their associated postures in patients with acute low back pain
. Spine 1995; 20: 2303–12.
13. Eisen EA. Methodology for analyzing episodic events. Scand J Work Environ Health 1999; 25: 36–42.
14. Feyer AM, Herbison P, Williamson AM, et al. The role of physical and psychological factors in occupational low back pain
: a prospective cohort study. Occup Environ Med 2000; 57: 116–20.
15. Frank JW, Pulcins IR, Kerr MS, et al. Occupational back pain: an unhelpful polemic. Scand J Work Environ Health 1995; 21: 3–14.
16. Garcy P, Mayer T, Gatchel RJ. Recurrent or new injury outcomes after return to work in chronic disabling spinal disorders: tertiary prevention efficacy of functional restoration treatment. Spine 1996; 21: 952–9.
17. Goertz MN. Prognostic indicators for acute low-back pain. Spine 1990; 15: 1307–10.
18. Heliovaara M, Sievers K, Impivaara O, et al. Descriptive epidemiology and public health aspects of low back pain
. Ann Med 1989; 21: 327–33.
19. Infante-Rivard C, Lortie M. Prognostic factors for return to work after a first compensated episode of back pain. Occup Environ Med 1996; 53: 488–94.
20. Infante-Rivard C, Lortie M. Relapse and short sickness absence for back pain in the six months after return to work. Occup Environ Med 1997; 54: 328–34.
21. Krause N, Dasinger LK, Deegan LJ, et al. Alternative approaches for measuring duration of work disability after low back injury based on administrative workers’ compensation data. Am J Ind Med 1999; 35: 604–18.
22. Macfarlane GJ, Thomas E, Croft PR, et al. Predictors of early improvement in low back pain
amongst consulters to general practice: the influence of pre-morbid and episode-related factors. Pain 1999; 80: 113–9.
23. Macfarlane GJ, Thomas E, Papageorgiou AC, et al. Employment and physical work activities as predictors of future low back pain
. Spine 1997; 22: 1143–9.
24. McGorry RW, Webster BS, Snook SH, et al. The relation between pain intensity, disability, and the episodic nature of chronic and recurrent low back pain
. Spine 2000; 25: 834–41.
25. McPhillips-Tangum CA, Cherkin DC, Rhodes LA, et al. Reasons for repeated medical visits among patients with chronic back pain. J Gen Intern Med 1998; 13: 289–95.
26. Miedema HS, Chorus AM, Wevers CW, et al. Chronicity of back problems during working life. Spine 1998; 23: 2021–9.
27. Miller JS, Pinnington MA, Stanley IM. The early stages of low back pain
: a pilot study of patient diaries as a source of data. Fam Pract 1999; 16: 395–401.
28. Nachemson A, Bigos SJ. The low back. In: Cruess JRW, ed. Adult Orthopedics. New York, NY: Churchill-Livingstone; 1984: 843–937.
29. Nachemson AL, Jonsson E. Neck and Back Pain: The Scientific Evidence of Causes, Diagnosis and Treatment. Philadelphia, PA: Lippincott Williams & Wilkins; 2000.
30. Oleinick A, Gluck JV, Guire K. Factors affecting first return to work following a compensable occupational back injury. Am J Ind Med 1996; 30: 540–55.
31. Papageorgiou AC, Croft PR, Thomas E, et al. Influence of previous pain experience on the episode incidence of low back pain
: results from the South Manchester Back Pain Study. Pain 1996; 66: 181–5.
32. Papageorgiou AC, Croft PR, Thomas E, et al. Psychosocial risks for low back pain
: are these related to work? Ann Rheum Dis 1998; 57: 500–2.
33. Papageorgiou AC, Macfarlane GJ, Thomas E, et al. Psychosocial factors in the workplace: do they predict new episodes
of low back pain
? Evidence from the South Manchester Back Pain Study. Spine 1997; 22: 1137–42.
34. Pope DP. Prevalence of shoulder pain in the community: the influence of case definition. Ann Rheum Dis 1997; 56: 308–12.
35. Reis S, Hermoni D, Borkan JM, et al. A new look at low back complaints in primary care
: a RAMBAM Israeli Family Practice Research Network study. J Fam Pract 1999; 48: 299–303.
36. Roland MO, Morrell DC, Morris RW. Can general practitioners predict the outcome of episodes
of back pain? Br Med J 1983; 286: 523–5.
37. Shekelle PG, Markovich M, Louie R. Comparing the costs between provider types of episodes
of back pain care
. Spine 1995; 20: 221–7.
38. Shekelle PG, Markovich M, Louie R. Factors associated with choosing a chiropractor for episodes
of back pain care
. Med Care
1995; 33: 842–50.
39. Shekelle PG, Markovich M, Louie R. An epidemiologic study of episodes
of back pain care
. Spine 1995; 20: 1668–73.
40. Skargren EI, Carlsson PG, Oberg BE. One-year follow-up comparison of the cost and effectiveness of chiropractic and physiotherapy as primary management for back pain: subgroup analysis, recurrence, and additional health care
utilization. Spine 1998; 23: 1875–84.
41. Smedley J, Inskip H, Cooper C, et al. Natural history of low back pain
: a longitudinal study in nurses. Spine 1998; 23: 2422–6.
42. Smith M, Stano M. Costs and recurrences of chiropractic and medical episodes
of low-back care
. J Manipulative Physiol Ther 1997; 20: 5–12.
43. Soukup MG, Glomsrod B, Lonn JH, et al. The effect of a Mensendieck exercise program as secondary prophylaxis for recurrent low back pain
: a randomized, controlled trial with 12-month follow-up. Spine 1999; 24: 1585–92.
44. Sundararajan V, Konrad TR, Garrett J, et al. Patterns and determinants of multiple provider use in patients with acute low back pain
. J Gen Intern Med 1998; 13: 528–33.
45. Von Korff M. Studying the natural history of back pain. Spine 1994; 19 (suppl): 2041–6.
46. Waddell G. A new clinical model for the treatment of low back pain
. Spine 1987; 22: 128–56.
47. Waddell G. The back pain revolution. Edinburgh: Churchill Livingstone, 1998.
48. Waxman R, Tennant A, Helliwell P. A prospective follow-up study of low back pain
in the community. Spine 2000; 25: 2085–90.