- Email: [email protected]
Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice
Manual Therapy xxx (2013) 1e7
Contents lists available at ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice Heidi Eirikstoft, Alice Kongsted* The Nordic Institute of Chiropractic and Clinical Biomechanics, Campusvej 55, 5230 Odense M, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 February 2013 Received in revised form 4 July 2013 Accepted 15 July 2013
Sub-grouping of low back pain (LBP) is believed to improve prediction of prognosis and treatment effects. The objectives of this study were: (1) to examine whether chiropractic patients could be sub-grouped according to an existing pathoanatomically-based classification system, (2) to describe patient characteristics within each subgroup, and (3) to determine the proportion of patients in whom clinicians considered the classification to be unchanged after approximately 10 days. A cohort of 923 LBP patients was included during their first consultation. Patients completed an extensive questionnaire and were examined according to a standardised protocol. Based on the clinical examination, patients were classified into diagnostic subgroups. After approximately 10 days, chiropractors reported whether they considered the subgroup had changed. The most frequent subgroups were reducible and partly reducible disc syndromes followed by facet joint pain, dysfunction and sacroiliac (SI)-joint pain. Classification was inconclusive in 5% of the patients. Differences in pain, activity limitation, and psychological factors were small across subgroups. Within 10 days, 82% were reported to belong to the same subgroup as at the first visit. In conclusion, LBP patients could be classified according to a standardised protocol, and chiropractors considered most patient classifications to be unchanged within 10 days. Differences in patient characteristics between subgroups were very small, and the clinical relevance of the classification system should be investigated by testing its value as a prognostic factor or a treatment effect modifier. It is recommended that this classification system be combined with psychological and social factors if it is to be useful. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Classification Cohort studies Low back pain Primary health care
1. Introduction Low back pain (LBP) is the cause of a high number of health care consultations, but provable treatment effects are modest and different treatments seem to have more or less the same effects (van Middelkoop et al., 2010; Rubinstein et al., 2011; Standaert et al., 2011). This has partly been attributed to the fact that randomised controlled trials often investigate the effect of a ‘one size fits all’ approach in which all patients with non-specific LBP have the same type of care, and it has been suggested that treatment effects may be improved by classification of non-specific LBP into homogeneous subgroups that can guide the choice of treatment
* Corresponding author. NIKKB, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark. Tel.: þ45 65504531. E-mail address: [email protected] (A. Kongsted).
(Leboeuf-Yde and Manniche, 2001; Kent and Keating, 2004; Hill et al., 2011). In 1987, the biopsychosocial model was suggested as a theoretical framework for the treatment of LBP (Waddell, 1987) and in the absence of specific diagnoses with consequences for outcome, profiling patients on the basis of biological, psychological and social prognostic factors appears relevant (Hemingway et al., 2013). Prognostic research has identified a high number of factors associated with outcome in LBP, but no single prognostic factor has been identified that strongly affects outcome in itself (Kent and Keating, 2008; Chou and Shekelle, 2010). Potentially relevant factors, and perhaps especially biological factors, are under-investigated in high quality studies (Kent and Keating, 2008; Hancock et al., 2011). To enhance the clinical usefulness of prognostic factors and treatment effect modifiers, a number of classification systems have been developed that combine such factors into predictive models or classification systems (Fairbank et al., 2011; Karayannis et al.,
1356-689X/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.math.2013.07.007
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
2
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
2012). A classification system developed by Petersen et al. was designed to subgroup LBP patients according to the most likely pathoanatomical diagnosis (Petersen et al., 2003). This system combines Mechanical Diagnosis and Therapy (MDT) (McKenzie and May, 2003) with tests for sacroiliac (SI)-joint pain, neurological signs, adverse neural tension tests, and non-organic signs. In that way, clinical findings are combined into classes that are potentially stronger biological components in a biopsychosocial model than single tests. The reliability of Petersen et al.’s system was tested in two small cohorts where the inter-tester agreement was found to be acceptable for the largest classes (Petersen et al., 2004; Kongsted and Leboeuf-Yde, 2010), whereas the reliability for the smaller diagnostic classes is unknown. Further, preliminary results suggest that the classification system has some predictive capacity (Kongsted and Leboeuf-Yde, 2010). However, its ability to categorise patients with similar profiles into the same subgroup, including its value as a prognostic factor or treatment effect modifier, has yet to be proven. As a basic step to explore the usefulness of the system developed by Petersen et al. and to investigate whether classification based on clinical findings results in classes that also differ on psychological and social factors, the objectives of this study were: (1) to examine whether chiropractic patients could be classified according to the classification system, (2) to describe patient characteristics within each subgroup, and (3) to determine the proportion of patients in whom clinicians considered the classification to be unchanged after approximately ten days. 2. Methods Patients with LBP who attended a clinic in the research network of the Nordic Institute for Chiropractic and Clinical Biomechanics in Denmark were recruited during their first visit for the current episode. Participants completed a questionnaire at baseline and the chiropractors classified patients based on a standardised examination protocol. Approximately 10 days after the initial visit, the clinicians registered whether they considered the diagnostic class had changed since baseline. The chiropractors were free to plan treatment that they deemed appropriate. The project was approved by the Danish Data Protection Agency (J-no. 7840-1011743), and the local ethics committee declared that the study did not need ethics approval according to Danish rules (DNcoBR, 2011). 2.1. Setting Thirty-six chiropractors from 17 clinics geographically spread across Denmark participated in the study. Prior to data collection, all clinicians participated in a one-day seminar covering the theory and practice of the examination protocol for the study. Particular focus was put on the MDT approach since this was the part of the protocol with which clinicians were the least familiar. After the seminar, clinicians were asked to practise the standardised examination and one of the authors (HE) visited all clinicians to train them in the execution of the project protocol. If a chiropractor did not seem conversant with the protocol, another visit was scheduled. Three chiropractors in one clinic withdrew from the project after this introduction because they found that the protocol differed too much from their usual clinical procedures. The clinical experience of the participating chiropractors varied from one to more than 20 years. 2.2. Study sample The chiropractors were instructed to include patients consecutively in the project as they contacted the clinic. Patients were
potential participants if they sought care because of LBP with or without leg pain, were aged 18e65 years, had access to a mobile phone and were able to use text messaging (because of follow-up procedures unrelated to the objectives of this study), and could read and understand Danish. Exclusion criteria were pregnancy, suspicion of inflammatory or pathological pain, acute referral to surgery, and having had more than one health care visit for LBP within the last three months. Prior to giving consent, oral and written information about the study procedures was delivered by the chiropractor or by a secretary. 2.3. Survey data Patients who gave consent to participate completed a questionnaire in the reception area before being examined by the chiropractor. The questionnaire was returned to the secretary in a sealed envelope and posted to the research unit. 2.3.1. Socio-demographics Socio-demographic factors were gender, age, physical work load (mainly sitting, sitting and walking, light physical work, hard physical work), and sick-leave (proportion reporting any days off work due to LBP within the previous month). 2.3.2. LBP characteristics Pain items were duration of the current episode (0e2 weeks, 2e 4 weeks, 1e3 months, >3 months), the number of previous episodes (0, 1e3, >3), the number of LBP days over the previous year (30 days, >30 days) (Hestbaek et al., 2003), LBP intensity (typical pain the previous week on a numeric rating scale (NRS) 0e10 (Dionne et al., 2008)), leg pain intensity (0e10 NRS typical pain last week), and leg pain (proportion with NRS >0). Activity limitation was measured using the Danish Roland Morris Disability Questionnaire (Albert et al., 2003) and summed as a proportional score (0e100) (Kent and Lauridsen, 2011). 2.3.3. Psychological factors Depressive symptoms were measured by the Major Depression Inventory (0e50) (Bech et al., 2001), pain-related fear of movement by the Fear Avoidance Beliefs Questionnaire (FABQ-work (0e42)); FABQ-physical activity (0e24) (Waddell et al., 1993), and coping by means of a single item from the Orebro Pain Questionnaire (Linton and Boersma, 2003) (‘Based on all the things you do to cope, or deal with your pain, on an average day, how much are you able to decrease it?’ 0 ¼ can’t decrease it at all; 10 ¼ can decrease it completely). FABQ-work was only asked of those who were working. 2.3.4. General health Self-reported general health was measured by the EQ-5D VAS (0 ¼ worst imaginable health state; 100 ¼ best imaginable health state) (Rabin and de Charro, 2001). 2.4. Subgroup classification The clinical examination included responses to repeated endrange movements (MDT testing), five pain provocation tests for SI-joint testing (Laslett, 2008), tests for adverse neural tension, Waddell’s non-organic signs (Waddell et al., 1980), and a neurological examination including straight leg raise (SLR) and tests of muscle strength, sensation and deep tendon reflexes. In addition to the test procedures, the protocol contained questions aimed at identifying signs of spinal stenosis and facet joint pain (Petersen et al., 2003). The examination findings were translated into diagnostic classes as suggested by Petersen et al., although in our study, we
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
allowed clinicians to register both a primary and a secondary diagnostic class and we added a class named ‘partly reducible disc’ that included patients who centralised partly on MDT testing but who did not meet criteria for a reducible disc. The diagnostic class referred to by Petersen et al. as ‘nerve root entrapment’ was included in the ‘adherent nerve root’ class as previously suggested (Petersen et al., 2004). Although the diagnostic classes were named according to the most likely pathoanatomical pain generator, these names should be considered labels rather than firm diagnostic descriptors. An overview of the diagnostic classes is listed in Appendix 1. Findings were registered in a web-based examination scheme and transferred to a central server. When the diagnostic class based on the predefined algorithm for classification was registered, chiropractors were asked whether they agreed with this diagnosis, based on their clinical judgement. Prevalence of SLR (classified as positive if radiating pain in the leg was provoked or aggravated) within the diagnostic classes was the only finding from the clinical examination that was included in the current description, since that finding has been shown to be associated with prognosis (Kent and Keating, 2008) and was not a finding that would lead to a particular classification on its own. 2.5. Subsequent assessment of the classification A web-based questionnaire was completed by the chiropractors approximately 10 days after the first visit or when the patient was seen for a second visit. The questions asked were “Based on the patient’s subsequent visit(s), is there reason to alter the classification made at the first visit?” (yes, no, not sure, have not seen this patient again), and “If yes, which diagnostic class does the patient belong to now?”. The answers were based on information and findings from a routine follow-up consultation and clinicians were not asked to repeat the entire standardised examination. 2.6. Data analysis No imputation was made for missing data, and classes with less than five patients were not described. This study was purely descriptive and statistical testing of differences across all diagnostic groups was not considered relevant. Observed potentially relevant
3
differences between a diagnostic class and the rest of the cohort were tested by means of a Chi-squared test (proportions), t-test (normal distributed continuous variables), or rank sum test (continuous variables with a non-normal distribution). Nominal variables were described as proportions with 95% confidence intervals (CI), and continuous variables as means with standard deviations (SD) or medians with inter-quartile ranges (IQR). Analyses were performed in STATA/SE 12.1. 3. Results A total of 951 patients were recruited between September 2010 and January 2012. Both clinician- and patient-reported data were available for 923 (Fig. 1) of these patients. Each of the 17 clinics recruited from 14 to 189 patients. Characteristics of the cohort are summarised in Table 1. The highest proportion of missing items on patient-reported variables was 5% (FABQ physical activity), and SLR was missing in 7%. 3.1. Diagnostic classification The most frequent diagnostic classes were reducible (24%), and partly reducible (24%), disc syndromes followed by facet joint pain (14%), dysfunction (11%) and SI-joint pain (8%) (Table 2). The diagnostic classes of adherent nerve root, non-mechanical disc syndrome, adverse neural tension, and spinal stenosis included less than five patients each and were not included in the description of patient characteristics. Five percent of patients were classified as inconclusive. Chiropractors agreed with the diagnostic classification in 66% of patients, were unsure or agreed partially in 20%, and did not agree in 11%. Chiropractors most frequently agreed with the classifications of facet joint pain and SI-joint pain, whereas they most often disagreed with the classes of ‘non-mechanical disc’ and ‘inconclusive’ (Table 3). 3.2. Patient characteristics in the diagnostic classes Patient characteristics in each diagnostic class appear in Table 2. Generally, differences were small in size and only some factors are
Fig. 1. Study flow-chart.
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
4
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
3.3. Subsequent assessment of the classification
Table 1 Descriptive data and number of missing items. Total cohort, n ¼ 923 Females, % (95% CI) Age in years, mean (SD) Physical work load, % (95% CI) Sitting Sitting & walking Light physical Heavy physical Sick-listing, % (95% CI) Duration, % (95% CI) 0e2 weeks 2e4 weeks 1e3 months >3 months LBP 0e10, median (IQR) Leg pain 0e10, median (IQR) Any leg pain, % (95% CI) Previous LBP episodes, % (95% CI) None 1e3 >3 >30 days LBP last year, % (95% CI) Activity limitation, 0e100 median (IQR) Depression, 0e50 median (IQR) FABQ-W, 0e42 median (IQR) FABQ-PA, 0e24 median (IQR) Coping, 0e10 mean (SD) Positive SLR, % (95% CI) a
45 (42e49) 43 (12) 23 35 20 22 23
(20e26) (32e39) (17e23) (19e25) (20e26)
63 14 10 13 7 2 58
(60e66) (11e16) (8e12) (11e16) (5e8) (0e4) (55e61)
16 35 49 26 52 6 11 13 5.9 11
(14e19) (32e38) (46e52) (23e29) (35e70) (3e11) (6e20) (9e16) (2.4) (9e14)
No. of missing items 0 0 3a
14a 20
An assessment of diagnostic class based on a routine follow-up consultation was completed for 715 (77%) of the participants. In 82%, the chiropractor reported that the diagnostic class had not changed, 8% that the diagnostic class had changed, 6% that it was ‘unclear’ whether the diagnosis had changed, and 4% of the patients were not seen since the first consultation. The diagnostic class in which the largest proportion was considered to have changed class was irreducible disc syndrome (19% changed). These 10 patients changed to partly reducible or reducible disc (n ¼ 3), nerve root compression (n ¼ 1), SI-joint pain (n ¼ 3), or facet joint pain (n ¼ 3). 4. Discussion
27 45 45 21
34 16 17 36a 48 28 69
Missing values among participants who were working.
commented on below. p-Values refer to a comparison of the mentioned diagnostic class with the rest of the cohort. 3.2.1. Socio-demographics A higher proportion of females was observed in SI-joint syndrome (p < 0.01). Overall, 22% of the cohort reported sick leave due to LBP or leg pain, whereas 57% of the patients with nerve root compression had been sick-listed. Patients with myofascial pain were the least likely to report any sick-listing (p ¼ 0.01). 3.2.2. LBP characteristics Most patients had experienced LBP previously and 49% reported more than three previous episodes. The majority of the patients had experienced LBP during this episode for less than two weeks. The duration of the current episode was longer for patients with postural syndrome (p < 0.05) or dysfunction (p < 0.01) when compared with all other classes (Table 2). The median LBP intensity was 7 and similar across classes, although patients with nerve root compression had less intensive LBP (p < 0.01). Leg pain intensity was lower than LBP intensity except for those patients with nerve root compression, who reported not only more intensive leg pain than LBP, but also more leg pain than other classes (p < 0.001). Activity limitation was most pronounced in irreducible and partly reducible disc syndromes, whereas patients with dysfunction or myofascial pain had the least limitations (all mentioned differences p < 0.001 as compared with the rest of the cohort) (Table 2). 3.2.3. Psychological factors Depression scores were low and with small variations between diagnostic classes (Table 2). Fear-avoidance beliefs concerning physical activity also did not differ significantly between groups, but fear-avoidance beliefs related to work were significantly more common with nerve root compression (p < 0.05).
Based on a standardised examination protocol, it was possible to classify the majority of LBP patients seeking care from chiropractors according to predefined criteria. However, in one third of the patients, clinicians stated that they did not fully agree with the diagnostic conclusion reached by the classification system. Furthermore, patient characteristics were so similar across the diagnostic classes that it is questionable whether the classification subgroups were clinically relevant. The most frequent diagnostic classes were reducible and partly reducible disc syndrome (47% of the cohort), which resembles the 46% classified with reducible disc syndrome when Petersen et al. used this classification system in a secondary care setting (Petersen et al., 2004). In a study we previously conducted in Danish chiropractic practice, only 27% were classified with disc syndromes (Kongsted and Leboeuf-Yde, 2010), whereas a number of studies using the MDT system classified around 80% with ‘derangement syndrome’ (another term for what we named disc syndromes) (Kilpikoski et al., 2002; Clare et al., 2005; May, 2006). In the current cohort, 14% were classified as facet joint syndrome and 11% as dysfunction, which were high proportions when compared with Petersen et al.’s secondary care cohort, in which only 1e3% of patients were classified into these categories. In the former chiropractic practice study, dysfunction was the most frequent class (32%) and 6% were classified as facet joint syndrome (Kongsted and Leboeuf-Yde, 2010). It may be that more patients in primary care fit into these categories, or another explanation could be that traditionally chiropractors have focused more on the facet joints and joint dysfunction as sources of back pain and this might influence their interpretation of test results. Differences in sampling between populations do not seem to be the main reason for these classification differences since our results were more similar to those from a secondary care setting than to the results of our previous study from chiropractic practices. Other possible explanations are differences in performing the clinical examination or in interpretations of test results. In our previous study, less emphasis was put on the hierarchy of classes and on MDT concepts. This meant that chiropractors in that study may have used end-range loading in a less rigorous manner. Chiropractors in the current study were taught by an MDT diploma therapist and a hierarchy of classes was emphasised more. Whether focus on MDT testing resulted in a more ‘correct’ classification or in clinicians being biased towards the disc syndromes is unknown. Comparing baseline characteristics across diagnostic classes did not reveal specific patient profiles belonging to each class, and observed differences between classes were small in size, which has been noted also with other classification approaches (Fritz and George, 2000; Kongsted et al., 2012). Nevertheless, some clinically meaningful differences were seen. First, it did make sense that
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
5
Table 2 Clinical characteristics of LBP patients in nine diagnostic classes.* Nerve root compression n ¼ 10
Myofascial pain n ¼ 42
35 (26;43) 46 (11)
70 (40;100) 50 (9)
45 (30;60) 44 (11)
(0;65) (0;42) (0;65) (0;65) (0;65)
24 29 28 19 18
(16;33) (20;38) (19;36) (12;27) (10;26)
29 (0;65) 29 (0;65) 43 (3;83) 57 (17;97)
24 29 21 26 5
(10;37) (14;44) (8;34) (12;41) (0;12)
(0;57) (0;37) (0;37) (13;87)
67 14 10 9
(59;75) (8;20) (5;16) (4;14)
25 25 25 25
50 17 14 19
(35;65) (5;28) (4;25) (7;31)
Reducible disc n ¼ 220
Partly reducible disc n ¼ 224
Irreducible disc n ¼ 60
Dysfunction n ¼ 98
SI-joint pain n ¼ 75
Postural syndrome n¼8
Facet joint pain n ¼ 127
43 (37;50) 43 (11)
48 (41;54) 41 (11)
52 (39;64) 43 (12)
41 (31;51) 42 (13)
65 (54;76) 45 (12)
38 (2;73) 44 (11)
20 39 22 19 25
(14;26) (32;46) (16;28) (13;25) (19;32)
34 33 16 18 29
(27;41) (26;39) (11;21) (12;23) (23;36)
16 48 16 20 18
(5;27) (33;63) (5;27) (8;33) (6;30)
14 41 17 28 14
(6;22) (30;52) (9;26) (17;38) (6;22)
19 37 21 23 24
(9;30) (24;49) (10;32) (12;34) (12;35)
29 14 29 29 29
Duration, % (95% CI) 0e2 weeks 2e4 weeks 1e3 months >3 months
67 12 12 8
(61;74) (8;16) (8;17) (5;12)
69 10 8 13
(63;76) (6;14) (5;12) (8;17)
65 15 8 12
(53;77) (6;24) (1;15) (3;20)
46 21 12 22
(36;56) (12;29) (6;19) (13;30)
64 13 8 15
(53;75) (5;20) (2;15) (7;24)
25 13 13 50
LBP 0e10, median (IQR) Leg pain 0e10, median (IQR) Any leg pain, % (95% CI)
7 (5;8) 2 (0;4) 62 (55;68)
7 (6;8) 2 (0;5) 59 (52;65)
7 (6;8) 2 (0;6) 63 (50;75)
6 (5;7) 1 (0;3) 64 (52;75)
7 (5;8) 2 (0;6) 62 (51;74)
5 (3;8) 1 (0;3) 63 (27;98)
7 (5;8) 0 (0;3) 44 (35;53)
3 (2;6) 6 (5;9) 100 (100;100)
6 (4;8) 2 (0;5) 65 (50;80)
Previous LBP episodes, % (95% CI) None 1e3 >3
14 (9;18) 36 (30;42) 50 (44;57)
15 (11;20) 34 (27;40) 51 (44;58)
10 (2;18) 39 (26;52) 51 (40;64)
21 (13;29) 33 (24;42) 46 (36;56)
21 (11;30) 19 (10;29) 60 (48;71)
13 (0;37) 38 (2;73) 50 (13;87)
17 (10;24) 40 (34;51) 40 (32;49)
13 (0;37) 50 (13;87) 38 (2;73)
14 (3;25) 36 (21;50) 50 (35;65)
24 (18;30) 57 (35;70)
27 (21;33) 61 (39;73)
28 (16;39) 65 (43;78)
29 (19;38) 39 (26;59)
27 (17;38) 53(39;74)
63 (27;98) 39 (8;50)
18 (11;25) 52 (30;68)
29 (0;65) 57 (39;74)
33 (19;49) 26 (17;61)
6 (3;11) 11 (6;21) 12 (9;16) 5.9 (2.4) 75(60;80)
6 (3;12) 10 (6;18) 13 (9;16) 5.9 (2.4) 70(60;80)
8 (4;14) 12 (5;21) 14 (10;18) 5.8 (2.3) 70(43;80)
5 (2;9) 10 (6;17) 13 (11;17) 5.9 (2.4) 70(60;85)
6 (3;12) 8 (4;19) 13 (9;17) 5.8 (2.5) 70(50;80)
8 (5;15) 18 (7;22) 12 (9;17) 6.7 (2.3) 70(60;85)
6 (3;12) 13 (7;19) 13 (9;17) 5.8 (2.4) 75(60;87)
8 (4;15) 23 (12;28) 14 (8;18) 5.4 (2.9) 75(48;86)
5 (3;8) 15 (6;21) 11 (7;15) 6.5 (2.4) 80(70;88)
14 (10;19)
10 (6;14)
23 (11;35)
2 (0;5)
9 (2;17)
14 (0;49)
2 (0;5)
Females, % (95% CI) Age in years, mean (SD) Physical work load, % (95% CI) Sitting Sitting& walking Light physical Heavy physical Any sick-listing, % (95% CI)
>30 days LBP last year, % (95% CI) Activity limitation, 0e100 median (IQR) Depression, 0e50 median (IQR) FABQ-W, 0e42 median (IQR) FABQ-PA, 0e24 median (IQR) Coping, 0e10 mean (SD) General Health, 0e100 median (IQR) Positive SLR, % (95% CI)
(0;57) (0;57) (0;57) (0;57)
80 (50;100)
5 (0;14)
*Due to few observations in the classes, non-mechanical disc (n ¼ 2), adherent nerve root (n ¼ 2), spinal stenosis (n ¼ 3), adverse neural tension (n ¼ 2) and abnormal pain syndrome (n ¼ 0) are not included in the table. The classification was inconclusive in 50 (5%) cases.
more women had SI-joint pain if some pelvic pain can be attributed to previous pregnancy and childbirth. Also, a more severe profile on both pain and psychological factors in patients with nerve root compression was in line with other studies (BenDebba et al., 2000; Kongsted et al., 2012). Finally, it seems reasonable from a clinical point of view that patients in the irreducible disc class (that is, movement tests worsen, not improve pain) tended to be more Table 3 Proportion of patients in each diagnostic class for whom clinicians did not agree that the conclusion reached by the classification system was correct. Diagnostic class
Clinician agrees with the conclusion
Clinician disagrees with the conclusion
Clinician unsure or agrees partly
Reducible disc n ¼ 220 Partly reducible disc n ¼ 224 Irreducible disc n ¼ 60 Non-mechanical disc n ¼ 2 Nerve root compression n ¼ 10 Postural syndrome n ¼ 8 SI-joint pain n ¼ 75 Dysfunction n ¼ 98 Adherent nerve root n ¼ 2 Facet joint pain n ¼ 127 Spinal stenosis n ¼ 3 Myofascial pain n ¼ 42 Adverse neural tension n ¼ 2 Inconclusive n ¼ 50
74% 69% 42% 50% 70% 38% 83% 64% 50% 83% 67% 38% 0% 12%
5% 11% 27% 50% 0% 25% 1% 9% 0% 2% 0% 26% 0% 36%
19% 18% 30% 0% 10% 38% 12% 23% 50% 13% 33% 24% 100% 40%
Because of missing responses regarding ‘agreement’ not all rows sum to 100%.
severely affected than those with a reducible disc. We could possibly have introduced clearer differences between classes if the class definitions had involved the patient history to a higher degree. We chose to define classes based almost purely on clinical examination findings in order to study the importance of the clinical examination separately from patient-reported information. It was a potential limitation of the study that clinicians were not formally trained in MDT, since MDT test procedures were a significant part of the study protocol. Also, standardisation of the test protocol was based only on a one-day course and one or two visits by a research assistant. On the other hand, all participating chiropractors were experienced in LBP examinations and the examination protocol resembled to a large extent a ‘standard’ LBP examination in Danish chiropractic practice. As mentioned above, it can also be argued that very strict standardisation could result in bias. However, it appears from the results that definitions were not always followed. For example, leg pain was present in patients classified with postural syndrome, which was defined as a syndrome with local LBP only. We believe inaccuracies like this are an almost inevitable outcome of conducting research in a ‘real life’ clinical setting involving a large number of data collectors. It is unknown to what extent the study population was representative of chiropractic patients. The participating clinicians were instructed to include new patients consecutively, but this was not always practically possible because extra time was needed to conduct the study procedures. Some patients calling for an appointment were not willing to wait until a ‘project appointment’
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
6
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
was available, and were therefore not included. It is our impression that lack of time was the main reason for not inviting a potential participant to become involved in the study, and we do not suspect that this influenced the representativeness of the cohort. In conclusion, LBP patients could be classified according to a predefined classification system on their first visit to a chiropractor. However, chiropractors often stated that they disagreed with the classification and this should be considered if future research suggests that the classification be implemented in practice. It is questionable whether diagnostic classes with such small differences in patient characteristics are of any relevance. Our results imply that this classification system would most likely not be an adequate description of patients in chiropractic care, but it is possible that it results in biological entities that
would be relevant in combination with psychological and social factors. The next step needed is to investigate the predictive value of the system, its usefulness in the identification of patients who respond to certain interventions, and its potential moderating effect on psychological and social factors. Acknowledgements The Danish ‘Chiropractic Fund’ financed the data collection and AK’s salary. The McKenzie Institute Denmark and The Danish Physiotherapist Association provided financial support for HE’s salary. We would like to thank the data collecting chiropractic clinics and research assistants Jytte Johannesen and Orla Lund Nielsen.
Appendix A. Summary of the definitions of diagnostic classes.
Diagnostic class
Definition
Reducible disc*
LBP leg pain Min. 1 painful movement direction, and Centralisation present (pain can move from one body region to another) LBP leg pain Pain tends to centralise but
Partly reducible disc
B Does not move from one region to another, or B Pain relief is not completely sustained, or B Local back pain is reduced but not abolished Irreducible disc*
Non-mechanical disc*
Nerve root compression
Spinal stenosis
Sacroiliac-joint syndrome
Dysfunction syndrome*
Postural syndrome*
Facet joint syndrome
LBP leg pain Min. 1 painful movement direction, and No centralisation, and Peripheralisation or increased pain with at least one movement direction LBP leg pain Criteria for other disc types not met, and pain gets worse after test of all movement directions, and No centralisation, and At least one of: Dominating pain above S1 Acute lateral shift Symptoms move from one side to the other when testing lateral glide or rotation Leg pain worse than LBP Positive straight leg raise (SLR) test or Reverse SLR, and Reduced muscle strength or impaired tendon reflex Criteria for disc or nerve root not met Symptoms improved with sitting or walking distance increased with lumbar flexion, and Best position is sitting or worst position is standing/walking LBP leg pain Does not meet criteria for disc or nerve root, and Min. any 3 of 5 SI-joint pain provocation tests positive SI tests: compression, distraction, sacral thrust, Gaenslen’ test, thigh thrust LBP pain Does not meet criteria for disc or nerve root, and Min. one movement direction restricted, and Pain is ONLY present at end-range of the restricted movement, and No change in symptoms following testing LBP Pain only present with static loading, and Normal range of motion, and No pain with movement tests, and Sustained static loading provokes well-known pain LBP Does not meet criteria for any of the above mentioned, and Pain with combined extension/rotation, and at least two of: Best position is sitting Best activity is walking Pain started in one side (lateral to the midline) Age >50 years
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
7
(continued ) Diagnostic class
Definition
Adverse neural tension (ANT)
LBP leg pain Only used as primary diagnostic class if criteria for the above-mentioned are not met Positive test for abnormal nerve tension Pain only present when stretching the painful tissue LBP leg pain Only used as primary diagnostic class if criteria for the above-mentioned other than ANT are not met Well-known pain is provoked by palpation of a tender muscle spot Only used as primary diagnostic class if criteria for other classes are not met At least 3 of: Widespread soreness LBP with axial compression or simulated rotation SLR improved when patient is distracted Reduced muscle strength or sensation in a non-anatomical pattern Vigorous pain reaction when examined Non-specific low back pain patients not included in any of the above listed classes
Myofascial pain
Abnormal pain syndrome
Inconclusive
*Reducible disc, irreducible disc and non-mechanical disc were defined the same way as derangement classes in to the Mechanical Diagnosis and Therapy classification (McKenzie and May, 2003), dysfunction syndrome and postural syndrome were adopted directly from the said classification system.
References Albert HB, Jensen AM, Dahl D, Rasmussen MN. Criteria validation of the Roland Morris questionnaire. A Danish translation of the international scale for the assessment of functional level in patients with low back pain and sciatica. Ugeskr Laeger 2003;165:1875e80. Bech P, Rasmussen NA, Olsen LR, Noerholm V, Abildgaard W. The sensitivity and specificity of the Major Depression Inventory, using the Present State Examination as the index of diagnostic validity. J Affect Disord 2001;66:159e64. BenDebba M, Torgerson WS, Long DM. A validated, practical classification procedure for many persistent low back pain patients. Pain 2000;87:89e97. Chou R, Shekelle P. Will this patient develop persistent disabling low back pain? JAMA 2010;303:1295e302. Clare HA, Adams R, Maher CG. Reliability of McKenzie classification of patients with cervical or lumbar pain. J Manipulative Physiol Ther 2005;28:122e7. Dionne CE, Dunn KM, Croft PR, Nachemson AL, Buchbinder R, Walker BF, et al. A consensus approach toward the standardization of back pain definitions for use in prevalence studies. Spine 2008;33:95e103. DNcoBR E, Danish National Committee on Biomedical Research Ethics. Guidelines about notification. http://www.cvk.sum.dk/English/guidelinesaboutnotifica tion.aspx; 2011. Fairbank J, Gwilym SE, France JC, Daffner SD, Dettori J, Hermsmeyer J, et al. The role of classification of chronic low back pain. Spine (Phila Pa 1976) 2011;36:S19e42. Fritz JM, George S. The use of a classification approach to identify subgroups of patients with acute low back pain. Interrater reliability and short-term treatment outcomes. Spine (Phila Pa 1976) 2000;25:106e14. Hancock MJ, Maher CG, Laslett M, Hay E, Koes B. Discussion paper: what happened to the ‘bio’ in the bio-psycho-social model of low back pain? Eur Spine J 2011;20:2105e10. Hemingway H, Croft P, Perel P, Hayden JA, Abrams K, Timmis A, et al. Prognosis research strategy (PROGRESS) 1: a framework for researching clinical outcomes. BMJ 2013;346:e5595. Hestbaek L, Leboeuf-Yde C, Engberg M, Lauritzen T, Bruun NH, Manniche C. The course of low back pain in a general population. Results from a 5-year prospective study. J Manipulative Physiol Ther 2003;26:213e9. Hill JC, Whitehurst DG, Lewis M, Bryan S, Dunn KM, Foster NE, et al. Comparison of stratified primary care management for low back pain with current best practice (STarT Back): a randomised controlled trial. Lancet 2011;378:1560e71. Karayannis NV, Jull GA, Hodges PW. Physiotherapy movement based classification approaches to low back pain: comparison of subgroups through review and developer/expert survey. BMC Musculoskelet Disord 2012;13:24. Kent P, Keating J. Do primary-care clinicians think that nonspecific low back pain is one condition? Spine 2004;29:1022e31. Kent P, Lauridsen HH. Managing missing scores on the Roland Morris Disability Questionnaire. Spine (Phila Pa 1976) 2011;36:1878e84.
Kent PM, Keating JL. Can we predict poor recovery from recent-onset nonspecific low back pain? A systematic review. Man Ther 2008;13:12e28. Kilpikoski S, Airaksinen O, Kankaanpaa M, Leminen P, Videman T, Alen M. Interexaminer reliability of low back pain assessment using the McKenzie method. Spine (Phila Pa 1976) 2002;27(8):E207e14. Kongsted A, Kent P, Albert HB, Jensen TS, Manniche C. Patients with low back pain differ from those who also have leg pain or signs of nerve root involvement e a cross-sectional study. BMC Musculoskelet Disord 2012:13. Kongsted A, Leboeuf-Yde C. The Nordic back pain subpopulation program: can low back pain patterns be predicted from the first consultation with a chiropractor? A longitudinal pilot study. Chiropr Osteopat 2010;18:8. Laslett M. Evidence-based diagnosis and treatment of the painful sacroiliac joint. J Man Manip Ther 2008;16:142e52. Leboeuf-Yde C, Manniche C. Low back pain: time to get off the treadmill. J Manipulative Physiol Ther 2001;24:63e6. Linton SJ, Boersma K. Early identification of patients at risk of developing a persistent back problem: the predictive validity of the Orebro Musculoskeletal Pain Questionnaire. Clin J Pain 2003;19:80e6. May S. Classification by McKenzie mechanical syndromes: a survey of McKenzietrained faculty. J Manipulative Physiol Ther 2006;29:637e42. McKenzie RA, May S. The lumbar spine: mechanical diagnosis & therapy; 2003. Petersen T, Laslett M, Thorsen H, Manniche C, Ekdahl C, Jacobsen S. Diagnostic classification of non-specific low back pain. A new system integrating pathoanatomic and clinical categories. Physiother Theory Pract 2003;19:213e37. Petersen T, Olsen S, Laslett M, Thorsen H, Manniche C, Ekdahl C, et al. Inter-tester reliability of a new diagnostic classification system for patients with nonspecific low back pain. Aust J Physiother 2004;50:85e94. Rabin R, de Charro F. EQ-5D: a measure of health status from the EuroQol Group. Ann Med 2001;33:337e43. Rubinstein SM, van MM, Assendelft WJ, de Boer MR, van Tulder MW. Spinal manipulative therapy for chronic low-back pain: an update of a Cochrane review. Spine (Phila Pa 1976) 2011;36(13):E825e46. Standaert CJ, Friedly J, Erwin MW, Lee MJ, Rechtine G, Henrikson NB, et al. Comparative effectiveness of exercise, acupuncture, and spinal manipulation for low back pain. Spine (Phila Pa 1976) 2011;36(Suppl. 21):S120e30. van Middelkoop M, Rubinstein SM, Verhagen AP, Ostelo RW, Koes BW, van Tulder MW. Exercise therapy for chronic nonspecific low-back pain. Best Pract Res Clin Rheumatol 2010;24:193e204. Waddell G. 1987 Volvo award in clinical sciences. A new clinical model for the treatment of low-back pain. Spine (Phila Pa 1976) 1987;12:632e44. Waddell G, McCulloch JA, Kummel E, Venner RM. Nonorganic physical signs in lowback pain. Spine (Phila Pa 1976) 1980;5:117e25. Waddell G, Newton M, Henderson I, Somerville D, Main CJ. A Fear-Avoidance Beliefs Questionnaire (FABQ) and the role of fear- avoidance beliefs in chronic low back pain and disability. Pain 1993;52:157e68.
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
Contents lists available at ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice Heidi Eirikstoft, Alice Kongsted* The Nordic Institute of Chiropractic and Clinical Biomechanics, Campusvej 55, 5230 Odense M, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 February 2013 Received in revised form 4 July 2013 Accepted 15 July 2013
Sub-grouping of low back pain (LBP) is believed to improve prediction of prognosis and treatment effects. The objectives of this study were: (1) to examine whether chiropractic patients could be sub-grouped according to an existing pathoanatomically-based classification system, (2) to describe patient characteristics within each subgroup, and (3) to determine the proportion of patients in whom clinicians considered the classification to be unchanged after approximately 10 days. A cohort of 923 LBP patients was included during their first consultation. Patients completed an extensive questionnaire and were examined according to a standardised protocol. Based on the clinical examination, patients were classified into diagnostic subgroups. After approximately 10 days, chiropractors reported whether they considered the subgroup had changed. The most frequent subgroups were reducible and partly reducible disc syndromes followed by facet joint pain, dysfunction and sacroiliac (SI)-joint pain. Classification was inconclusive in 5% of the patients. Differences in pain, activity limitation, and psychological factors were small across subgroups. Within 10 days, 82% were reported to belong to the same subgroup as at the first visit. In conclusion, LBP patients could be classified according to a standardised protocol, and chiropractors considered most patient classifications to be unchanged within 10 days. Differences in patient characteristics between subgroups were very small, and the clinical relevance of the classification system should be investigated by testing its value as a prognostic factor or a treatment effect modifier. It is recommended that this classification system be combined with psychological and social factors if it is to be useful. Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Classification Cohort studies Low back pain Primary health care
1. Introduction Low back pain (LBP) is the cause of a high number of health care consultations, but provable treatment effects are modest and different treatments seem to have more or less the same effects (van Middelkoop et al., 2010; Rubinstein et al., 2011; Standaert et al., 2011). This has partly been attributed to the fact that randomised controlled trials often investigate the effect of a ‘one size fits all’ approach in which all patients with non-specific LBP have the same type of care, and it has been suggested that treatment effects may be improved by classification of non-specific LBP into homogeneous subgroups that can guide the choice of treatment
* Corresponding author. NIKKB, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark. Tel.: þ45 65504531. E-mail address: [email protected] (A. Kongsted).
(Leboeuf-Yde and Manniche, 2001; Kent and Keating, 2004; Hill et al., 2011). In 1987, the biopsychosocial model was suggested as a theoretical framework for the treatment of LBP (Waddell, 1987) and in the absence of specific diagnoses with consequences for outcome, profiling patients on the basis of biological, psychological and social prognostic factors appears relevant (Hemingway et al., 2013). Prognostic research has identified a high number of factors associated with outcome in LBP, but no single prognostic factor has been identified that strongly affects outcome in itself (Kent and Keating, 2008; Chou and Shekelle, 2010). Potentially relevant factors, and perhaps especially biological factors, are under-investigated in high quality studies (Kent and Keating, 2008; Hancock et al., 2011). To enhance the clinical usefulness of prognostic factors and treatment effect modifiers, a number of classification systems have been developed that combine such factors into predictive models or classification systems (Fairbank et al., 2011; Karayannis et al.,
1356-689X/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.math.2013.07.007
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
2
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
2012). A classification system developed by Petersen et al. was designed to subgroup LBP patients according to the most likely pathoanatomical diagnosis (Petersen et al., 2003). This system combines Mechanical Diagnosis and Therapy (MDT) (McKenzie and May, 2003) with tests for sacroiliac (SI)-joint pain, neurological signs, adverse neural tension tests, and non-organic signs. In that way, clinical findings are combined into classes that are potentially stronger biological components in a biopsychosocial model than single tests. The reliability of Petersen et al.’s system was tested in two small cohorts where the inter-tester agreement was found to be acceptable for the largest classes (Petersen et al., 2004; Kongsted and Leboeuf-Yde, 2010), whereas the reliability for the smaller diagnostic classes is unknown. Further, preliminary results suggest that the classification system has some predictive capacity (Kongsted and Leboeuf-Yde, 2010). However, its ability to categorise patients with similar profiles into the same subgroup, including its value as a prognostic factor or treatment effect modifier, has yet to be proven. As a basic step to explore the usefulness of the system developed by Petersen et al. and to investigate whether classification based on clinical findings results in classes that also differ on psychological and social factors, the objectives of this study were: (1) to examine whether chiropractic patients could be classified according to the classification system, (2) to describe patient characteristics within each subgroup, and (3) to determine the proportion of patients in whom clinicians considered the classification to be unchanged after approximately ten days. 2. Methods Patients with LBP who attended a clinic in the research network of the Nordic Institute for Chiropractic and Clinical Biomechanics in Denmark were recruited during their first visit for the current episode. Participants completed a questionnaire at baseline and the chiropractors classified patients based on a standardised examination protocol. Approximately 10 days after the initial visit, the clinicians registered whether they considered the diagnostic class had changed since baseline. The chiropractors were free to plan treatment that they deemed appropriate. The project was approved by the Danish Data Protection Agency (J-no. 7840-1011743), and the local ethics committee declared that the study did not need ethics approval according to Danish rules (DNcoBR, 2011). 2.1. Setting Thirty-six chiropractors from 17 clinics geographically spread across Denmark participated in the study. Prior to data collection, all clinicians participated in a one-day seminar covering the theory and practice of the examination protocol for the study. Particular focus was put on the MDT approach since this was the part of the protocol with which clinicians were the least familiar. After the seminar, clinicians were asked to practise the standardised examination and one of the authors (HE) visited all clinicians to train them in the execution of the project protocol. If a chiropractor did not seem conversant with the protocol, another visit was scheduled. Three chiropractors in one clinic withdrew from the project after this introduction because they found that the protocol differed too much from their usual clinical procedures. The clinical experience of the participating chiropractors varied from one to more than 20 years. 2.2. Study sample The chiropractors were instructed to include patients consecutively in the project as they contacted the clinic. Patients were
potential participants if they sought care because of LBP with or without leg pain, were aged 18e65 years, had access to a mobile phone and were able to use text messaging (because of follow-up procedures unrelated to the objectives of this study), and could read and understand Danish. Exclusion criteria were pregnancy, suspicion of inflammatory or pathological pain, acute referral to surgery, and having had more than one health care visit for LBP within the last three months. Prior to giving consent, oral and written information about the study procedures was delivered by the chiropractor or by a secretary. 2.3. Survey data Patients who gave consent to participate completed a questionnaire in the reception area before being examined by the chiropractor. The questionnaire was returned to the secretary in a sealed envelope and posted to the research unit. 2.3.1. Socio-demographics Socio-demographic factors were gender, age, physical work load (mainly sitting, sitting and walking, light physical work, hard physical work), and sick-leave (proportion reporting any days off work due to LBP within the previous month). 2.3.2. LBP characteristics Pain items were duration of the current episode (0e2 weeks, 2e 4 weeks, 1e3 months, >3 months), the number of previous episodes (0, 1e3, >3), the number of LBP days over the previous year (30 days, >30 days) (Hestbaek et al., 2003), LBP intensity (typical pain the previous week on a numeric rating scale (NRS) 0e10 (Dionne et al., 2008)), leg pain intensity (0e10 NRS typical pain last week), and leg pain (proportion with NRS >0). Activity limitation was measured using the Danish Roland Morris Disability Questionnaire (Albert et al., 2003) and summed as a proportional score (0e100) (Kent and Lauridsen, 2011). 2.3.3. Psychological factors Depressive symptoms were measured by the Major Depression Inventory (0e50) (Bech et al., 2001), pain-related fear of movement by the Fear Avoidance Beliefs Questionnaire (FABQ-work (0e42)); FABQ-physical activity (0e24) (Waddell et al., 1993), and coping by means of a single item from the Orebro Pain Questionnaire (Linton and Boersma, 2003) (‘Based on all the things you do to cope, or deal with your pain, on an average day, how much are you able to decrease it?’ 0 ¼ can’t decrease it at all; 10 ¼ can decrease it completely). FABQ-work was only asked of those who were working. 2.3.4. General health Self-reported general health was measured by the EQ-5D VAS (0 ¼ worst imaginable health state; 100 ¼ best imaginable health state) (Rabin and de Charro, 2001). 2.4. Subgroup classification The clinical examination included responses to repeated endrange movements (MDT testing), five pain provocation tests for SI-joint testing (Laslett, 2008), tests for adverse neural tension, Waddell’s non-organic signs (Waddell et al., 1980), and a neurological examination including straight leg raise (SLR) and tests of muscle strength, sensation and deep tendon reflexes. In addition to the test procedures, the protocol contained questions aimed at identifying signs of spinal stenosis and facet joint pain (Petersen et al., 2003). The examination findings were translated into diagnostic classes as suggested by Petersen et al., although in our study, we
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
allowed clinicians to register both a primary and a secondary diagnostic class and we added a class named ‘partly reducible disc’ that included patients who centralised partly on MDT testing but who did not meet criteria for a reducible disc. The diagnostic class referred to by Petersen et al. as ‘nerve root entrapment’ was included in the ‘adherent nerve root’ class as previously suggested (Petersen et al., 2004). Although the diagnostic classes were named according to the most likely pathoanatomical pain generator, these names should be considered labels rather than firm diagnostic descriptors. An overview of the diagnostic classes is listed in Appendix 1. Findings were registered in a web-based examination scheme and transferred to a central server. When the diagnostic class based on the predefined algorithm for classification was registered, chiropractors were asked whether they agreed with this diagnosis, based on their clinical judgement. Prevalence of SLR (classified as positive if radiating pain in the leg was provoked or aggravated) within the diagnostic classes was the only finding from the clinical examination that was included in the current description, since that finding has been shown to be associated with prognosis (Kent and Keating, 2008) and was not a finding that would lead to a particular classification on its own. 2.5. Subsequent assessment of the classification A web-based questionnaire was completed by the chiropractors approximately 10 days after the first visit or when the patient was seen for a second visit. The questions asked were “Based on the patient’s subsequent visit(s), is there reason to alter the classification made at the first visit?” (yes, no, not sure, have not seen this patient again), and “If yes, which diagnostic class does the patient belong to now?”. The answers were based on information and findings from a routine follow-up consultation and clinicians were not asked to repeat the entire standardised examination. 2.6. Data analysis No imputation was made for missing data, and classes with less than five patients were not described. This study was purely descriptive and statistical testing of differences across all diagnostic groups was not considered relevant. Observed potentially relevant
3
differences between a diagnostic class and the rest of the cohort were tested by means of a Chi-squared test (proportions), t-test (normal distributed continuous variables), or rank sum test (continuous variables with a non-normal distribution). Nominal variables were described as proportions with 95% confidence intervals (CI), and continuous variables as means with standard deviations (SD) or medians with inter-quartile ranges (IQR). Analyses were performed in STATA/SE 12.1. 3. Results A total of 951 patients were recruited between September 2010 and January 2012. Both clinician- and patient-reported data were available for 923 (Fig. 1) of these patients. Each of the 17 clinics recruited from 14 to 189 patients. Characteristics of the cohort are summarised in Table 1. The highest proportion of missing items on patient-reported variables was 5% (FABQ physical activity), and SLR was missing in 7%. 3.1. Diagnostic classification The most frequent diagnostic classes were reducible (24%), and partly reducible (24%), disc syndromes followed by facet joint pain (14%), dysfunction (11%) and SI-joint pain (8%) (Table 2). The diagnostic classes of adherent nerve root, non-mechanical disc syndrome, adverse neural tension, and spinal stenosis included less than five patients each and were not included in the description of patient characteristics. Five percent of patients were classified as inconclusive. Chiropractors agreed with the diagnostic classification in 66% of patients, were unsure or agreed partially in 20%, and did not agree in 11%. Chiropractors most frequently agreed with the classifications of facet joint pain and SI-joint pain, whereas they most often disagreed with the classes of ‘non-mechanical disc’ and ‘inconclusive’ (Table 3). 3.2. Patient characteristics in the diagnostic classes Patient characteristics in each diagnostic class appear in Table 2. Generally, differences were small in size and only some factors are
Fig. 1. Study flow-chart.
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
4
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
3.3. Subsequent assessment of the classification
Table 1 Descriptive data and number of missing items. Total cohort, n ¼ 923 Females, % (95% CI) Age in years, mean (SD) Physical work load, % (95% CI) Sitting Sitting & walking Light physical Heavy physical Sick-listing, % (95% CI) Duration, % (95% CI) 0e2 weeks 2e4 weeks 1e3 months >3 months LBP 0e10, median (IQR) Leg pain 0e10, median (IQR) Any leg pain, % (95% CI) Previous LBP episodes, % (95% CI) None 1e3 >3 >30 days LBP last year, % (95% CI) Activity limitation, 0e100 median (IQR) Depression, 0e50 median (IQR) FABQ-W, 0e42 median (IQR) FABQ-PA, 0e24 median (IQR) Coping, 0e10 mean (SD) Positive SLR, % (95% CI) a
45 (42e49) 43 (12) 23 35 20 22 23
(20e26) (32e39) (17e23) (19e25) (20e26)
63 14 10 13 7 2 58
(60e66) (11e16) (8e12) (11e16) (5e8) (0e4) (55e61)
16 35 49 26 52 6 11 13 5.9 11
(14e19) (32e38) (46e52) (23e29) (35e70) (3e11) (6e20) (9e16) (2.4) (9e14)
No. of missing items 0 0 3a
14a 20
An assessment of diagnostic class based on a routine follow-up consultation was completed for 715 (77%) of the participants. In 82%, the chiropractor reported that the diagnostic class had not changed, 8% that the diagnostic class had changed, 6% that it was ‘unclear’ whether the diagnosis had changed, and 4% of the patients were not seen since the first consultation. The diagnostic class in which the largest proportion was considered to have changed class was irreducible disc syndrome (19% changed). These 10 patients changed to partly reducible or reducible disc (n ¼ 3), nerve root compression (n ¼ 1), SI-joint pain (n ¼ 3), or facet joint pain (n ¼ 3). 4. Discussion
27 45 45 21
34 16 17 36a 48 28 69
Missing values among participants who were working.
commented on below. p-Values refer to a comparison of the mentioned diagnostic class with the rest of the cohort. 3.2.1. Socio-demographics A higher proportion of females was observed in SI-joint syndrome (p < 0.01). Overall, 22% of the cohort reported sick leave due to LBP or leg pain, whereas 57% of the patients with nerve root compression had been sick-listed. Patients with myofascial pain were the least likely to report any sick-listing (p ¼ 0.01). 3.2.2. LBP characteristics Most patients had experienced LBP previously and 49% reported more than three previous episodes. The majority of the patients had experienced LBP during this episode for less than two weeks. The duration of the current episode was longer for patients with postural syndrome (p < 0.05) or dysfunction (p < 0.01) when compared with all other classes (Table 2). The median LBP intensity was 7 and similar across classes, although patients with nerve root compression had less intensive LBP (p < 0.01). Leg pain intensity was lower than LBP intensity except for those patients with nerve root compression, who reported not only more intensive leg pain than LBP, but also more leg pain than other classes (p < 0.001). Activity limitation was most pronounced in irreducible and partly reducible disc syndromes, whereas patients with dysfunction or myofascial pain had the least limitations (all mentioned differences p < 0.001 as compared with the rest of the cohort) (Table 2). 3.2.3. Psychological factors Depression scores were low and with small variations between diagnostic classes (Table 2). Fear-avoidance beliefs concerning physical activity also did not differ significantly between groups, but fear-avoidance beliefs related to work were significantly more common with nerve root compression (p < 0.05).
Based on a standardised examination protocol, it was possible to classify the majority of LBP patients seeking care from chiropractors according to predefined criteria. However, in one third of the patients, clinicians stated that they did not fully agree with the diagnostic conclusion reached by the classification system. Furthermore, patient characteristics were so similar across the diagnostic classes that it is questionable whether the classification subgroups were clinically relevant. The most frequent diagnostic classes were reducible and partly reducible disc syndrome (47% of the cohort), which resembles the 46% classified with reducible disc syndrome when Petersen et al. used this classification system in a secondary care setting (Petersen et al., 2004). In a study we previously conducted in Danish chiropractic practice, only 27% were classified with disc syndromes (Kongsted and Leboeuf-Yde, 2010), whereas a number of studies using the MDT system classified around 80% with ‘derangement syndrome’ (another term for what we named disc syndromes) (Kilpikoski et al., 2002; Clare et al., 2005; May, 2006). In the current cohort, 14% were classified as facet joint syndrome and 11% as dysfunction, which were high proportions when compared with Petersen et al.’s secondary care cohort, in which only 1e3% of patients were classified into these categories. In the former chiropractic practice study, dysfunction was the most frequent class (32%) and 6% were classified as facet joint syndrome (Kongsted and Leboeuf-Yde, 2010). It may be that more patients in primary care fit into these categories, or another explanation could be that traditionally chiropractors have focused more on the facet joints and joint dysfunction as sources of back pain and this might influence their interpretation of test results. Differences in sampling between populations do not seem to be the main reason for these classification differences since our results were more similar to those from a secondary care setting than to the results of our previous study from chiropractic practices. Other possible explanations are differences in performing the clinical examination or in interpretations of test results. In our previous study, less emphasis was put on the hierarchy of classes and on MDT concepts. This meant that chiropractors in that study may have used end-range loading in a less rigorous manner. Chiropractors in the current study were taught by an MDT diploma therapist and a hierarchy of classes was emphasised more. Whether focus on MDT testing resulted in a more ‘correct’ classification or in clinicians being biased towards the disc syndromes is unknown. Comparing baseline characteristics across diagnostic classes did not reveal specific patient profiles belonging to each class, and observed differences between classes were small in size, which has been noted also with other classification approaches (Fritz and George, 2000; Kongsted et al., 2012). Nevertheless, some clinically meaningful differences were seen. First, it did make sense that
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
5
Table 2 Clinical characteristics of LBP patients in nine diagnostic classes.* Nerve root compression n ¼ 10
Myofascial pain n ¼ 42
35 (26;43) 46 (11)
70 (40;100) 50 (9)
45 (30;60) 44 (11)
(0;65) (0;42) (0;65) (0;65) (0;65)
24 29 28 19 18
(16;33) (20;38) (19;36) (12;27) (10;26)
29 (0;65) 29 (0;65) 43 (3;83) 57 (17;97)
24 29 21 26 5
(10;37) (14;44) (8;34) (12;41) (0;12)
(0;57) (0;37) (0;37) (13;87)
67 14 10 9
(59;75) (8;20) (5;16) (4;14)
25 25 25 25
50 17 14 19
(35;65) (5;28) (4;25) (7;31)
Reducible disc n ¼ 220
Partly reducible disc n ¼ 224
Irreducible disc n ¼ 60
Dysfunction n ¼ 98
SI-joint pain n ¼ 75
Postural syndrome n¼8
Facet joint pain n ¼ 127
43 (37;50) 43 (11)
48 (41;54) 41 (11)
52 (39;64) 43 (12)
41 (31;51) 42 (13)
65 (54;76) 45 (12)
38 (2;73) 44 (11)
20 39 22 19 25
(14;26) (32;46) (16;28) (13;25) (19;32)
34 33 16 18 29
(27;41) (26;39) (11;21) (12;23) (23;36)
16 48 16 20 18
(5;27) (33;63) (5;27) (8;33) (6;30)
14 41 17 28 14
(6;22) (30;52) (9;26) (17;38) (6;22)
19 37 21 23 24
(9;30) (24;49) (10;32) (12;34) (12;35)
29 14 29 29 29
Duration, % (95% CI) 0e2 weeks 2e4 weeks 1e3 months >3 months
67 12 12 8
(61;74) (8;16) (8;17) (5;12)
69 10 8 13
(63;76) (6;14) (5;12) (8;17)
65 15 8 12
(53;77) (6;24) (1;15) (3;20)
46 21 12 22
(36;56) (12;29) (6;19) (13;30)
64 13 8 15
(53;75) (5;20) (2;15) (7;24)
25 13 13 50
LBP 0e10, median (IQR) Leg pain 0e10, median (IQR) Any leg pain, % (95% CI)
7 (5;8) 2 (0;4) 62 (55;68)
7 (6;8) 2 (0;5) 59 (52;65)
7 (6;8) 2 (0;6) 63 (50;75)
6 (5;7) 1 (0;3) 64 (52;75)
7 (5;8) 2 (0;6) 62 (51;74)
5 (3;8) 1 (0;3) 63 (27;98)
7 (5;8) 0 (0;3) 44 (35;53)
3 (2;6) 6 (5;9) 100 (100;100)
6 (4;8) 2 (0;5) 65 (50;80)
Previous LBP episodes, % (95% CI) None 1e3 >3
14 (9;18) 36 (30;42) 50 (44;57)
15 (11;20) 34 (27;40) 51 (44;58)
10 (2;18) 39 (26;52) 51 (40;64)
21 (13;29) 33 (24;42) 46 (36;56)
21 (11;30) 19 (10;29) 60 (48;71)
13 (0;37) 38 (2;73) 50 (13;87)
17 (10;24) 40 (34;51) 40 (32;49)
13 (0;37) 50 (13;87) 38 (2;73)
14 (3;25) 36 (21;50) 50 (35;65)
24 (18;30) 57 (35;70)
27 (21;33) 61 (39;73)
28 (16;39) 65 (43;78)
29 (19;38) 39 (26;59)
27 (17;38) 53(39;74)
63 (27;98) 39 (8;50)
18 (11;25) 52 (30;68)
29 (0;65) 57 (39;74)
33 (19;49) 26 (17;61)
6 (3;11) 11 (6;21) 12 (9;16) 5.9 (2.4) 75(60;80)
6 (3;12) 10 (6;18) 13 (9;16) 5.9 (2.4) 70(60;80)
8 (4;14) 12 (5;21) 14 (10;18) 5.8 (2.3) 70(43;80)
5 (2;9) 10 (6;17) 13 (11;17) 5.9 (2.4) 70(60;85)
6 (3;12) 8 (4;19) 13 (9;17) 5.8 (2.5) 70(50;80)
8 (5;15) 18 (7;22) 12 (9;17) 6.7 (2.3) 70(60;85)
6 (3;12) 13 (7;19) 13 (9;17) 5.8 (2.4) 75(60;87)
8 (4;15) 23 (12;28) 14 (8;18) 5.4 (2.9) 75(48;86)
5 (3;8) 15 (6;21) 11 (7;15) 6.5 (2.4) 80(70;88)
14 (10;19)
10 (6;14)
23 (11;35)
2 (0;5)
9 (2;17)
14 (0;49)
2 (0;5)
Females, % (95% CI) Age in years, mean (SD) Physical work load, % (95% CI) Sitting Sitting& walking Light physical Heavy physical Any sick-listing, % (95% CI)
>30 days LBP last year, % (95% CI) Activity limitation, 0e100 median (IQR) Depression, 0e50 median (IQR) FABQ-W, 0e42 median (IQR) FABQ-PA, 0e24 median (IQR) Coping, 0e10 mean (SD) General Health, 0e100 median (IQR) Positive SLR, % (95% CI)
(0;57) (0;57) (0;57) (0;57)
80 (50;100)
5 (0;14)
*Due to few observations in the classes, non-mechanical disc (n ¼ 2), adherent nerve root (n ¼ 2), spinal stenosis (n ¼ 3), adverse neural tension (n ¼ 2) and abnormal pain syndrome (n ¼ 0) are not included in the table. The classification was inconclusive in 50 (5%) cases.
more women had SI-joint pain if some pelvic pain can be attributed to previous pregnancy and childbirth. Also, a more severe profile on both pain and psychological factors in patients with nerve root compression was in line with other studies (BenDebba et al., 2000; Kongsted et al., 2012). Finally, it seems reasonable from a clinical point of view that patients in the irreducible disc class (that is, movement tests worsen, not improve pain) tended to be more Table 3 Proportion of patients in each diagnostic class for whom clinicians did not agree that the conclusion reached by the classification system was correct. Diagnostic class
Clinician agrees with the conclusion
Clinician disagrees with the conclusion
Clinician unsure or agrees partly
Reducible disc n ¼ 220 Partly reducible disc n ¼ 224 Irreducible disc n ¼ 60 Non-mechanical disc n ¼ 2 Nerve root compression n ¼ 10 Postural syndrome n ¼ 8 SI-joint pain n ¼ 75 Dysfunction n ¼ 98 Adherent nerve root n ¼ 2 Facet joint pain n ¼ 127 Spinal stenosis n ¼ 3 Myofascial pain n ¼ 42 Adverse neural tension n ¼ 2 Inconclusive n ¼ 50
74% 69% 42% 50% 70% 38% 83% 64% 50% 83% 67% 38% 0% 12%
5% 11% 27% 50% 0% 25% 1% 9% 0% 2% 0% 26% 0% 36%
19% 18% 30% 0% 10% 38% 12% 23% 50% 13% 33% 24% 100% 40%
Because of missing responses regarding ‘agreement’ not all rows sum to 100%.
severely affected than those with a reducible disc. We could possibly have introduced clearer differences between classes if the class definitions had involved the patient history to a higher degree. We chose to define classes based almost purely on clinical examination findings in order to study the importance of the clinical examination separately from patient-reported information. It was a potential limitation of the study that clinicians were not formally trained in MDT, since MDT test procedures were a significant part of the study protocol. Also, standardisation of the test protocol was based only on a one-day course and one or two visits by a research assistant. On the other hand, all participating chiropractors were experienced in LBP examinations and the examination protocol resembled to a large extent a ‘standard’ LBP examination in Danish chiropractic practice. As mentioned above, it can also be argued that very strict standardisation could result in bias. However, it appears from the results that definitions were not always followed. For example, leg pain was present in patients classified with postural syndrome, which was defined as a syndrome with local LBP only. We believe inaccuracies like this are an almost inevitable outcome of conducting research in a ‘real life’ clinical setting involving a large number of data collectors. It is unknown to what extent the study population was representative of chiropractic patients. The participating clinicians were instructed to include new patients consecutively, but this was not always practically possible because extra time was needed to conduct the study procedures. Some patients calling for an appointment were not willing to wait until a ‘project appointment’
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
6
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
was available, and were therefore not included. It is our impression that lack of time was the main reason for not inviting a potential participant to become involved in the study, and we do not suspect that this influenced the representativeness of the cohort. In conclusion, LBP patients could be classified according to a predefined classification system on their first visit to a chiropractor. However, chiropractors often stated that they disagreed with the classification and this should be considered if future research suggests that the classification be implemented in practice. It is questionable whether diagnostic classes with such small differences in patient characteristics are of any relevance. Our results imply that this classification system would most likely not be an adequate description of patients in chiropractic care, but it is possible that it results in biological entities that
would be relevant in combination with psychological and social factors. The next step needed is to investigate the predictive value of the system, its usefulness in the identification of patients who respond to certain interventions, and its potential moderating effect on psychological and social factors. Acknowledgements The Danish ‘Chiropractic Fund’ financed the data collection and AK’s salary. The McKenzie Institute Denmark and The Danish Physiotherapist Association provided financial support for HE’s salary. We would like to thank the data collecting chiropractic clinics and research assistants Jytte Johannesen and Orla Lund Nielsen.
Appendix A. Summary of the definitions of diagnostic classes.
Diagnostic class
Definition
Reducible disc*
LBP leg pain Min. 1 painful movement direction, and Centralisation present (pain can move from one body region to another) LBP leg pain Pain tends to centralise but
Partly reducible disc
B Does not move from one region to another, or B Pain relief is not completely sustained, or B Local back pain is reduced but not abolished Irreducible disc*
Non-mechanical disc*
Nerve root compression
Spinal stenosis
Sacroiliac-joint syndrome
Dysfunction syndrome*
Postural syndrome*
Facet joint syndrome
LBP leg pain Min. 1 painful movement direction, and No centralisation, and Peripheralisation or increased pain with at least one movement direction LBP leg pain Criteria for other disc types not met, and pain gets worse after test of all movement directions, and No centralisation, and At least one of: Dominating pain above S1 Acute lateral shift Symptoms move from one side to the other when testing lateral glide or rotation Leg pain worse than LBP Positive straight leg raise (SLR) test or Reverse SLR, and Reduced muscle strength or impaired tendon reflex Criteria for disc or nerve root not met Symptoms improved with sitting or walking distance increased with lumbar flexion, and Best position is sitting or worst position is standing/walking LBP leg pain Does not meet criteria for disc or nerve root, and Min. any 3 of 5 SI-joint pain provocation tests positive SI tests: compression, distraction, sacral thrust, Gaenslen’ test, thigh thrust LBP pain Does not meet criteria for disc or nerve root, and Min. one movement direction restricted, and Pain is ONLY present at end-range of the restricted movement, and No change in symptoms following testing LBP Pain only present with static loading, and Normal range of motion, and No pain with movement tests, and Sustained static loading provokes well-known pain LBP Does not meet criteria for any of the above mentioned, and Pain with combined extension/rotation, and at least two of: Best position is sitting Best activity is walking Pain started in one side (lateral to the midline) Age >50 years
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007
H. Eirikstoft, A. Kongsted / Manual Therapy xxx (2013) 1e7
7
(continued ) Diagnostic class
Definition
Adverse neural tension (ANT)
LBP leg pain Only used as primary diagnostic class if criteria for the above-mentioned are not met Positive test for abnormal nerve tension Pain only present when stretching the painful tissue LBP leg pain Only used as primary diagnostic class if criteria for the above-mentioned other than ANT are not met Well-known pain is provoked by palpation of a tender muscle spot Only used as primary diagnostic class if criteria for other classes are not met At least 3 of: Widespread soreness LBP with axial compression or simulated rotation SLR improved when patient is distracted Reduced muscle strength or sensation in a non-anatomical pattern Vigorous pain reaction when examined Non-specific low back pain patients not included in any of the above listed classes
Myofascial pain
Abnormal pain syndrome
Inconclusive
*Reducible disc, irreducible disc and non-mechanical disc were defined the same way as derangement classes in to the Mechanical Diagnosis and Therapy classification (McKenzie and May, 2003), dysfunction syndrome and postural syndrome were adopted directly from the said classification system.
References Albert HB, Jensen AM, Dahl D, Rasmussen MN. Criteria validation of the Roland Morris questionnaire. A Danish translation of the international scale for the assessment of functional level in patients with low back pain and sciatica. Ugeskr Laeger 2003;165:1875e80. Bech P, Rasmussen NA, Olsen LR, Noerholm V, Abildgaard W. The sensitivity and specificity of the Major Depression Inventory, using the Present State Examination as the index of diagnostic validity. J Affect Disord 2001;66:159e64. BenDebba M, Torgerson WS, Long DM. A validated, practical classification procedure for many persistent low back pain patients. Pain 2000;87:89e97. Chou R, Shekelle P. Will this patient develop persistent disabling low back pain? JAMA 2010;303:1295e302. Clare HA, Adams R, Maher CG. Reliability of McKenzie classification of patients with cervical or lumbar pain. J Manipulative Physiol Ther 2005;28:122e7. Dionne CE, Dunn KM, Croft PR, Nachemson AL, Buchbinder R, Walker BF, et al. A consensus approach toward the standardization of back pain definitions for use in prevalence studies. Spine 2008;33:95e103. DNcoBR E, Danish National Committee on Biomedical Research Ethics. Guidelines about notification. http://www.cvk.sum.dk/English/guidelinesaboutnotifica tion.aspx; 2011. Fairbank J, Gwilym SE, France JC, Daffner SD, Dettori J, Hermsmeyer J, et al. The role of classification of chronic low back pain. Spine (Phila Pa 1976) 2011;36:S19e42. Fritz JM, George S. The use of a classification approach to identify subgroups of patients with acute low back pain. Interrater reliability and short-term treatment outcomes. Spine (Phila Pa 1976) 2000;25:106e14. Hancock MJ, Maher CG, Laslett M, Hay E, Koes B. Discussion paper: what happened to the ‘bio’ in the bio-psycho-social model of low back pain? Eur Spine J 2011;20:2105e10. Hemingway H, Croft P, Perel P, Hayden JA, Abrams K, Timmis A, et al. Prognosis research strategy (PROGRESS) 1: a framework for researching clinical outcomes. BMJ 2013;346:e5595. Hestbaek L, Leboeuf-Yde C, Engberg M, Lauritzen T, Bruun NH, Manniche C. The course of low back pain in a general population. Results from a 5-year prospective study. J Manipulative Physiol Ther 2003;26:213e9. Hill JC, Whitehurst DG, Lewis M, Bryan S, Dunn KM, Foster NE, et al. Comparison of stratified primary care management for low back pain with current best practice (STarT Back): a randomised controlled trial. Lancet 2011;378:1560e71. Karayannis NV, Jull GA, Hodges PW. Physiotherapy movement based classification approaches to low back pain: comparison of subgroups through review and developer/expert survey. BMC Musculoskelet Disord 2012;13:24. Kent P, Keating J. Do primary-care clinicians think that nonspecific low back pain is one condition? Spine 2004;29:1022e31. Kent P, Lauridsen HH. Managing missing scores on the Roland Morris Disability Questionnaire. Spine (Phila Pa 1976) 2011;36:1878e84.
Kent PM, Keating JL. Can we predict poor recovery from recent-onset nonspecific low back pain? A systematic review. Man Ther 2008;13:12e28. Kilpikoski S, Airaksinen O, Kankaanpaa M, Leminen P, Videman T, Alen M. Interexaminer reliability of low back pain assessment using the McKenzie method. Spine (Phila Pa 1976) 2002;27(8):E207e14. Kongsted A, Kent P, Albert HB, Jensen TS, Manniche C. Patients with low back pain differ from those who also have leg pain or signs of nerve root involvement e a cross-sectional study. BMC Musculoskelet Disord 2012:13. Kongsted A, Leboeuf-Yde C. The Nordic back pain subpopulation program: can low back pain patterns be predicted from the first consultation with a chiropractor? A longitudinal pilot study. Chiropr Osteopat 2010;18:8. Laslett M. Evidence-based diagnosis and treatment of the painful sacroiliac joint. J Man Manip Ther 2008;16:142e52. Leboeuf-Yde C, Manniche C. Low back pain: time to get off the treadmill. J Manipulative Physiol Ther 2001;24:63e6. Linton SJ, Boersma K. Early identification of patients at risk of developing a persistent back problem: the predictive validity of the Orebro Musculoskeletal Pain Questionnaire. Clin J Pain 2003;19:80e6. May S. Classification by McKenzie mechanical syndromes: a survey of McKenzietrained faculty. J Manipulative Physiol Ther 2006;29:637e42. McKenzie RA, May S. The lumbar spine: mechanical diagnosis & therapy; 2003. Petersen T, Laslett M, Thorsen H, Manniche C, Ekdahl C, Jacobsen S. Diagnostic classification of non-specific low back pain. A new system integrating pathoanatomic and clinical categories. Physiother Theory Pract 2003;19:213e37. Petersen T, Olsen S, Laslett M, Thorsen H, Manniche C, Ekdahl C, et al. Inter-tester reliability of a new diagnostic classification system for patients with nonspecific low back pain. Aust J Physiother 2004;50:85e94. Rabin R, de Charro F. EQ-5D: a measure of health status from the EuroQol Group. Ann Med 2001;33:337e43. Rubinstein SM, van MM, Assendelft WJ, de Boer MR, van Tulder MW. Spinal manipulative therapy for chronic low-back pain: an update of a Cochrane review. Spine (Phila Pa 1976) 2011;36(13):E825e46. Standaert CJ, Friedly J, Erwin MW, Lee MJ, Rechtine G, Henrikson NB, et al. Comparative effectiveness of exercise, acupuncture, and spinal manipulation for low back pain. Spine (Phila Pa 1976) 2011;36(Suppl. 21):S120e30. van Middelkoop M, Rubinstein SM, Verhagen AP, Ostelo RW, Koes BW, van Tulder MW. Exercise therapy for chronic nonspecific low-back pain. Best Pract Res Clin Rheumatol 2010;24:193e204. Waddell G. 1987 Volvo award in clinical sciences. A new clinical model for the treatment of low-back pain. Spine (Phila Pa 1976) 1987;12:632e44. Waddell G, McCulloch JA, Kummel E, Venner RM. Nonorganic physical signs in lowback pain. Spine (Phila Pa 1976) 1980;5:117e25. Waddell G, Newton M, Henderson I, Somerville D, Main CJ. A Fear-Avoidance Beliefs Questionnaire (FABQ) and the role of fear- avoidance beliefs in chronic low back pain and disability. Pain 1993;52:157e68.
Please cite this article in press as: Eirikstoft H, Kongsted A, Patient characteristics in low back pain subgroups based on an existing classification system. A descriptive cohort study in chiropractic practice, Manual Therapy (2013), http://dx.doi.org/10.1016/j.math.2013.07.007