FROM: J Manipulative Physiol Ther. 2005 (Sep); 27 (7): 472–478 ~ FULL TEXT
Charlotte Leboeuf-Yde, DC, MPH, Iben Axén, DC, Jess James Jones, MSc,
Annika Rosenbaum, BAppSc (Chiro), Peter W. Løvgren, DC,
Laszlo Halasz, MHSc (ClinBiomech), Kristian Larsen, MPH, PT
The Back Research Center,
Backcenter Funen and University of Southern Denmark,
DK-5750 Ringe, Denmark.
OBJECTIVES: (1) To describe the low back pain (LBP) pattern at baseline;
(2) to describe the long-term outcome pattern;
(3) to investigate the presence of distinct subgroups in relation to outcome;
(4) to establish whether short-term outcome is a predictor of long-term outcome.
METHODS: A 3- to 6- and 12- to 18-month, multicenter practice-based, prospective descriptive study was performed in private chiropractic practices in Sweden. Fifty-eight of 64 previously compliant chiropractors each recruited a maximum of 30 consecutive patients with LBP. Complete baseline clinical information was provided on 1054 patients, of which 93% were interviewed approximately 3 months later, and 57% responded to a questionnaire at approximately 12 months. Chiropractic treatment was decided by the treating chiropractor. Twelve descriptive subgroups were created based on
(1) duration of LBP at baseline,
(2) duration of LBP in the past year, and
(3) LBP pattern in the past year. The predictive value was tested for outcome status at the fourth visit.
Information on self-reported LBP status and improvement over the past months were collected.
RESULTS: Patients were spread in a U-shaped fashion from benign to severe with the 2 extreme groups being most prevalent. About half the participants reported "no LBP in the past week" at 3 months and somewhat fewer at 12 months. Almost 75% claimed to be definitely better at 3 months, and approximately 50% at 12 months. Specific predictive subgroups can be identified, mainly in relation to the past-year history of LBP. Improvement at the fourth visit is a predictor of long-term outcome.
CONCLUSION: Knowledge of specific subgroups may improve the quality of care and the selection of homogeneous study populations in clinical trials.
From the FULL TEXT Article:
There is a need to study the amorphous group consisting of patients with “nonspecific” low back pain (LBP) to find out if it consists of specific subgroups. If such specific subgroups exist, they may cause confusion when they, if unidentified, are grouped together and investigated in relation to cause and treatment.
Also, the long-term outcome pattern is not well understood in patients with LBP receiving chiropractic treatment. For these reasons, a prospective study was designed, in which a large number of patients typical to chiropractic practice were followed for more than 1 year, to find out whether it is possible to identify specific subgroups on the basis of their baseline characteristics and long-term outcome pattern. In addition, the prospective design of the study made it possible to consider whether outcomes at 3 and 12 months were associated with the short-term outcome.
The latter aspect is an important clinical issue, as patients and chiropractors often have divergent expectations of the treatment process. The patient expects it to be quick and will want to stop treatment as soon as the acute pain has abated, whereas the chiropractor wants to complete the treatment until the patient has reached his/her optimal status and is left with a feeling of frustration when patients disappear “prematurely.” It is not known which approach is the best, the “quick-fix” or the “let's-improve-the-spinal-function-to-its-maximum,” and for which types of patients.
It was previously noted that duration of LBP at the time of consultation is a predictor of short-term treatment outcome, so that patients with a maximum duration of 1 week were more likely to improve than those with a duration of 1 to 2 weeks,  and that the short-term prognosis for patients with a baseline duration of more than 2 weeks is even less favorable. 
In the present study, data were available on patients examined at baseline, interviewed after 3 months, and surveyed through a questionnaire after 12 months. This made it possible to obtain answers to a number of specific questions:
(1) What is the distribution of subgroups in chiropractic patients with LBP?
(2) What is the long-term outcome in chiropractic patients with LBP?
(3) Is the outcome pattern similar across the different subgroups and outcome variables?
(4) Is early improvement a predictor for long-term outcome in a mixed LBP group of chiropractic patients?
The present study, which consists of a large sample of patients obtained in a wide variety of chiropractic practices, seems to be the first study in which the long-term outcome is mapped in relation to various outcome measures and different baseline subpopulations after chiropractic treatment.
The response rate at the 3-month interview was excellent, and although only 57% returned their questionnaires in the 12-month postal survey, there appeared to be no obvious differences between responders and nonresponders in relation to age, sex, LBP characteristics at baseline, and the proportion of patients who reported definite improvement by the fourth visit.
In previous studies, 2 predefined study populations were investigated in relation to outcome pattern and predictors for outcome, namely, patients with
(1) less than 2 weeks of LBP at the time of the first consultation and nonpersistent LBP , and
(2) more than 2 weeks of LBP at the first consultation and persistent LBP in the preceding year. [2, 5]
It was interesting to note that these were 2 substantial subgroups in our total study population. In fact, chiropractic patients with LBP range from the 2 extremes (short-lasting, nonpersistent LBP of intermittent type vs more long-lasting, persistent LBP of daily type), with some of the other subgroups being fairly substantial.
Three months after the initial consultation, approximately half of the patients have no prevalent pain, and the majority report only minor (if any) problems in relation to their daily activities. The majority also felt generally better 3 months after the study than at baseline. Nevertheless, it is less common to claim shorter spells of LBP or reduced frequency of attacks, and even less common to feel that the back is more resistant to LBP.
As time passes by, at 12 months, the present status is considered somewhat less favorable, and the general improvement is rated somewhat lower. However, more patients report to be more resistant, to experience less frequent pain, and to have shorter duration of their LBP events. In other words, at 3 months patients feel good, but it is difficult to pinpoint specific areas of improvement. At 12 months, their LBP status is considered less favorable, but they, nevertheless, feel improved. It is not known whether this seemingly conflicting pattern of reporting can be trusted. However, it seems likely that patients can accurately describe whether they had LBP the last week or not, that they can remember if they were in some way disabled because of LBP, and that they can give an overall judgment on whether they have improved or not. However, it is perhaps more doubtful whether patients can accurately remember the level of discomfort and details of previous LBP for as long ago as 1 year. Also, as LBP is a recurrent condition for many,  the outcome pattern for these 3 variables (better results at 3 months than at 12 months) seems most plausible. Based on the face validity of these outcome variables (Table 3), it appears acceptable to use and trust the 2 “present-status” variables and also one of the “development-over-the-past-months” variables, namely, the most global one (“definitely better”).
It has been suggested that one reason for the relatively small mean differences in treatment outcome between different treatment methods could be the presence of underlying subgroups of patients, resulting in varying treatment effects.  According to this study (Table 4A, Table 4B), not only specific subgroups at baseline but also the choice of outcome variables will affect the results. Again, 2 outcome variables that can be used to detect differences between baseline subgroups at both 3 and 12 months are “no LBP past week” and “definitely better.”
As for the underlying baseline factors, an interesting pattern was noted:
The duration of the patient's present complaint (the LBP event) appears to have less importance than the pain pattern in the preceding year, which, in turn, is less important than the total duration of pain in the past year (the LBP disease). This concept, which is illustrated in Table 5A, Table 5B, is probably new to those clinicians who concentrate on the present LBP event when making their diagnosis, choice of treatment, and prognosis. In clinical practice, it will now be necessary to take into account that the prognoses at both 3 and 12 months are associated with the LBP disease.
In summary, our findings indicate that there are several distinct subgroups of chiropractic patients with LBP, based on a simple classification system relating to the LBP event and the LBP disease. Some of these subgroups are fairly substantial and can be used to predict treatment outcome in chiropractic practice as well as to create homogeneous study populations, for example, when testing treatment effect in clinical trials. Clinicians have a particular responsibility toward patients who have not shown considerable improvement by the fourth visit, as they are also more likely than others not to report improvement at 3 and 12 months.
It is noteworthy that the duration of patients' complaint at baseline (the circumstances surrounding the LBP event) is of less prognostic value than the duration in total over the past year (the circumstances surrounding the LBP disease).
Another important point is that attention should be paid not only to the history of LBP in the past year but also to the status at the fourth visit. Patients with persistent LBP with or without leg pain over the past year and those who have not reported definite improvement by the fourth visit will require particular vigilance to avoid long and unsuccesful treatment programs.
From a researcher's point of view, several distinct subgroups have been identified based on this simple classification system, and some of these are of sufficient size to make data collection feasible. Future randomized controlled clinical trials should concentrate on such specific subgroups. Care should also be taken in selecting appropriate outcome variables at different points in time.
Knowledge of specific subgroups may improve the quality of care and the selection of homogeneous study populations in clinical trials. The present study design should be considered to systematically test other classification systems with the view of identifying clinically relevant subgroups of patients with LBP.
Axen I, Rosenbaum A, Robech R, Larsen K, Leboeuf-Yde C.
The Nordic Back Pain Subpopulation Program: Can Patient Reactions
to the First Chiropractic Treatment Predict Early Favorable
Treatment Outcome in Nonpersistent Low Back Pain?
J Manipulative Physiol Ther. 2005 (Mar); 28 (3): 153–158
Axen I, Rosenbaum A, Robech R, Larsen K, Leboeuf-Yde C.
Can Patient Reactions to the First Chiropractic Treatment Predict
Early Favorable Treatment Outcome in Nonpersistent Low Back Pain?
J Manipulative Physiol Ther. 2002 (Sep); 25 (7): 450–454
Axen I, Jones JJ, Rosenbaum A, Lovgren PW, Halasz L, Larsen K, et al.
The Nordic Back Pain Subpopulation Program: Validation and Improvement
of a Predictive Model for Treatment Outcome in Patients With
Low Back Pain Receiving Chiropractic Treatment
J Manipulative Physiol Ther. 2005 (Jul); 28 (6): 381–385
Leboeuf-Yde, C, Hennius, B, Rudberg, E, Leufvenmark, P, and Thunman, M.
Side effects of chiropractic treatment: a prospective study.
J Manipulative Physiol Ther. 1997; 20: 511–515
Leboeuf-Yde C, Gronstvedt A, Borge JA, Lothe J, Magnesen E. et al.
The Nordic Back Pain Subpopulation Program: Demographic and Clinical
Predictors for Outcome in Patients Receiving Chiropractic
Treatment for Persistent Low–Back Pain
J Manipulative Physiol Ther. 2004 (Oct); 27 (8): 493–502
Hestbæk, L, Leboeuf-Yde, C, and Manniche, C.
Low back pain—what is the long term course? A review of studies of general patient populations.
Eur Spine J. 2003; 11: 149–165
Leboeuf-Yde, C and Manniche, C.
Low back pain. Time to get off the treadmill.
J Manipulative Physiol Ther. 2001; 24: 53–55