Cynthia K. Peterson, DC, MMedEd, Jennifer Bolton, PhD, MAEd, B. Kim Humphreys, DC, PhD
Faculty of Medicine,
University of Zürich, Zürich, Switzerland.
OBJECTIVES: The purpose of this study was to investigate outcomes and prognostic factors in patients with acute or chronic low back pain (LBP) undergoing chiropractic treatment.
METHODS: This was a prognostic cohort study with medium-term outcomes. Adult patients with LBP of any duration who had not received chiropractic or manual therapy in the prior 3 months were recruited from multiple chiropractic practices in Switzerland. Participating doctors of chiropractic were allowed to use their typical treatment methods (such as chiropractic manipulation, soft tissue mobilization, or other methods) because the purpose of the study was to evaluate outcomes from routine chiropractic practice. Patients completed a numerical pain rating scale and Oswestry disability questionnaire immediately before treatment and at 1 week, 1 month, and 3 months after the start of treatment, together with self-reported improvement using the Patient Global Impression of Change.
RESULTS: Patients with acute (<4 weeks; n = 523) and chronic (>3 months; n = 293) LBP were included. Baseline mean pain and disability scores were significantly (P < .001) higher in patients with acute LBP. In both groups of patients, there were significant (P < .0001) improvements in mean scores of pain and disability at 1 week, 1 month, and 3 months, although these change scores were significantly greater in the acute group. Similarly, a greater proportion of patients in the acute group reported improvement at each follow-up. The most consistent predictor was self-reported improvement at 1 week, which was independently associated with improvement at 1 month (adjusted odds ratio [OR], 2.4 [95% confidence interval, 1.3-4.5] and 5.0 [2.4-10.6]) and at 3 months (2.9 [1.3-6.6] and 3.3 [1.3-8.7]) in patients with acute and chronic pain, respectively. The presence of radiculopathy at baseline was not a predictor of outcome.
CONCLUSIONS: Patients with chronic and acute pain reporting that they were “much better” or “better” on the Patient Global Impression of Change scale at 1 week after the first chiropractic visit were 4 to 5 times more likely to be improved at both 1 and 3 months compared with patients who were not improved at 1 week. Patients with acute pain reported more severe pain and disability initially but recovered faster. Patients with chronic and acute back pain both reported good outcomes, and most patients with radiculopathy also improved.
From the FULL TEXT Article
Research into mechanical low back pain (LBP) is ongoing because it is so common and leads to high health care costs. [1-19] The point prevalence of LBP is stated to be between 15% and 30%, the 1-year period prevalence between 15% and 45%, and a life-time prevalence of 50% to 80%. [7, 18, 19] It is also expensive to treat, with billions of dollars and billions of Euros spent per year in Western societies. [7, 18, 19] Between the years 1997 and 2005, the total estimated expenditures for patients with spine problems in the United States increased 65%, which was higher than overall health expenditures.  People with neck or back pain who also reported that physical impairment increased from 20.7% to 24.7% in that same period.  Low back pain is a significant burden not only to the individual who has it but also to their families, workplace, and society in general. In Europe, the yearly burden of LBP to society amounts to €211 per person in Sweden and €260 per person in the United Kingdom. 
Although LBP remains the most common musculoskeletal complaint presenting to physicians and other therapists, controversy remains surrounding the precise cause of the pain in many patients, appropriate therapies, subgroups of patients likely to improve with treatment and/or have more favorable prognoses, and associated medical costs. [1-10] To date, there is more evidence supporting chiropractic treatment of chronic LBP and LBP without concomitant radiating leg pain than for patients with acute or subacute pain and those with radiculopathy or sciatica. [2, 10-12] In chronic LBP, recent studies indicate that significant improvement is often fairly rapid, usually by the fourth visit, and that patients initially receiving treatment 3 to 4 times a week have better outcomes. [7, 8, 14] Although evidence is less clear for patients complaining of LBP of less than 3-month duration, it is suggested that spinal manipulation produces slightly better results than placebo, no treatment, and other noninvasive therapies. 
Some authors claim that patients more likely to respond to spinal manipulative therapy in the short term can be identified by specific physical and demographic factors.  However, many of these studies on spinal manipulative therapy are compromised by fairly small sample sizes and short-term follow-up, [2, 4, 7] and more research is needed to substantiate or refute these findings. Other larger studies with similar research protocols to this study have attempted to subgroup patients with LBP based on the clinical course of their condition. [8, 9, 13-15] Few demographic variables that are linked to clinical improvement in patients with LBP have been found thus far. The strongest and most consistent finding in studies looking at predictors of improvement is how quickly the patient responds to treatment. [8, 9, 13-15] It appears that if a patient will have a favorable outcome, the response to therapy is rapid. Although it is known that most patients with acute LBP improve very quickly and that this improvement may depend little on the treatment given, it is the patients with chronic LBP that provide the major challenge to clinicians and insurers. Patients with chronic LBP are also responsible for considerable costs because of a higher number of diagnostic procedures, treatment, and loss of productivity. [16, 17] Further large, prospective, cohort studies from other cultural environments are needed to confirm the prognostic factors reported in previous studies, as well as to investigate additional predictors of positive or negative outcomes, particularly coexisting radiculopathy.
It is not known if these same prognostic factors are applicable to chiropractic patients from other cultural or geographic regions. Therefore, this study was designed to investigate outcomes and prognostic factors linked with clinically significant improvement for patients with either acute or chronic LBP undergoing chiropractic treatment.
This prognostic cohort study with medium-term outcomes supports previous research that patients with LBP undergoing chiropractic treatment who are likely to respond do so very quickly. [8, 9, 13-15] This was not only true for patients presenting with acute (<4 weeks) LBP but, importantly, also for patients with chronic pain for greater than 3 months. Patients with chronic and acute pain reporting that they were much better or better on the PGIC scale at 1 week after the first chiropractic visit were 4 to 5 times more likely to be improved at both 1 and 3 months compared with patients who were not improved at 1 week. Patients who reported that they were “slightly better” on the PGIC scale were not considered improved in this study to err on the side of caution in defining clinically relevant improvement. Improvement is much more than just changes in the pain severity. It incorporates functional status and quality of life parameters, as well. The PGIC scale allows patients to express their multidimensional experience from their own viewpoint as to what is important to them in terms of “improvement” or “not-improved.”  Several other related studies, particularly from Denmark, have also reported that chiropractic patients who respond early on in the course of treatment usually have a more favorable outcome. [8-10, 13-15, 24] This current study, however, found that unlike previous work that reported that patients only improved up to week 7,  the patients with acute and chronic pain in Switzerland continued to improve on all outcome measures at each time point including at 3 months after the first chiropractic consultation.
Although in the univariate analysis, several potential predictors of outcome were identified, in the adjusted models, relatively few were identified as independently associated with improvement. None of the baseline variables collected in this study independently predicted outcome at 1 or 3 months for patients with acute or chronic pain, with the exception of trauma as the cause of the LBP in patients with chronic pain. This was only independently associated with outcome at 1 month, however. The most consistent independent association with improvement for both patients with acute pain and patients with chronic pain at both the 1- and 3-month time points was patient-reported improvement at 1 week after the start of treatment. Patients with acute and chronic pain who were improved at 1 week were approximately 3 times more likely to be improved at 3 months. Patients with chronic pain who also had a larger baseline to 1-month NRS change score were more likely to be improved at 3 months. Improvement at 1 month was also independently associated with improvement at 3 months for patients with acute pain.
The patients with acute pain in this study had significantly higher baseline NRS and Oswestry scores compared with the patients with chronic pain, as well as improved faster and with greater magnitude. At 3 months, 88% reported being much better or better. Their 3-month mean Oswestry score decreased by 74%, and their mean NRS score was 72% lower at 3 months compared with baseline. Much of this improvement is likely to be due to the natural history of LBP, and any specific treatment effects cannot be determined from a prospective cohort study of this type. However, even the patients with chronic pain demonstrated significant improvement at each time point, with 69% stating that they were much better or better at 3 months. This is unlikely to be due to the natural history of LBP because these patients have already passed the period when natural history occurs. The mean 3-month Oswestry score for the patients with chronic pain was decreased by 37%, and their mean NRS score was 39% lower than the baseline score. These values are well beyond the 30% reduction in NRS and Oswestry scores that has been identified as clinically meaningful improvement. [8, 25]
Although the Oswestry pain and disability questionnaire is probably not the best outcome measure for this patient population, having been designed for more severe and surgical cases, it was used because it had been translated and validated into German and French. Certainly, the fairly low mean baseline scores of 14.7 for patients with acute pain and 10.1 for patients with chronic pain of a maximum of 50 points reflects the inadequacy of this questionnaire for this patient population. Another questionnaire would have been preferred but was not available in the required languages. Interestingly, the univariate predictors showed that male patients were more likely to improve compared with female patients at 1 week, 1 month, and even 3 months in the acute patient group. Although sex was no longer a predictor of outcome in the multivariate model, other research has found that men have better outcomes than did women when undergoing chiropractic treatment. [9, 26] This sex difference was also seen when looking at the proportion of men and women in the acute and chronic subgroups in this study. Patients with acute pain showed a fairly even sex distribution, whereas a significantly higher percentage of patients with chronic pain were female (58%). This supports the observation that being male leads to a more favorable outcome when presenting with acute LBP. One possible explanation for the higher prevalence of patients with chronic pain being female is that degenerative lumbar instability is more common in middle-aged and older women. [27, 28]
An important and unique finding in this current study is that although 123 (23%) of the patients with acute LBP and 71 (24%) of the patients with chronic LBP were diagnosed by their chiropractors as having radiculopathy, this finding was not a negative predictor of improvement. Radiculopathy was not simply defined as leg pain but required clinical signs of nerve root compression as determined by the examining chiropractor. Previous studies investigating outcomes from patients with LBP undergoing spinal manipulation have purposely excluded patients with radiculopathy, [2, 10, 29] and others have found that the presence of leg pain is a negative predictor of improvement. [12, 24, 30] This study purposely included these patients to evaluate this subgroup. It is quite common for patients with LBP experiencing radiculopathy to seek chiropractic care in Switzerland and to receive spinal manipulative therapy as one of the treatment options.
The results of this study could help practicing chiropractors make more confident decisions about patient prognosis based on how quickly individual patients respond to their treatment. Chiropractors can also expect most of their patients with acute and chronic pain to continue to improve at least up to 3 months after the start of treatment, even if they are no longer being treated. These results also offer support for the use of chiropractic treatment for patients with radiculopathy.
Future follow-up studies from this project, in addition to the longer-term outcomes and prognostic variables being currently collected, could include qualitative studies comparing practice characteristics and techniques that may be linked with improved outcomes. During data entry, one of the authors noticed that certain practitioners seemed to obtain better patient outcomes compared with others. Obtaining detailed information concerning practice environments, types of treatments used, treatment schedules, doctor-patient communication styles, and types of patients who seek care from those individual practitioners with better patient outcomes would be beneficial for chiropractic education at both the undergraduate and the postgraduate level. These data could help to inform “best practice.”
Because this was a prospective cohort study and not a randomized trial, the observed outcomes cannot be definitively attributed to treatment. However, that a high percentage of patients with chronic pain improved is unlikely to be the result of the natural history of LBP. In addition, based on the pragmatic nature of this study, we do not know how many times the patients were treated during the treatment period or the specifics of the therapeutic interventions. However, based on the Swiss Job Analysis Survey 2009, it is likely that most patients were treated with diversified chiropractic spinal manipulation. It has been suggested that there is a positive relationship between the number of chiropractic treatments on LBP and disability for patients with chronic pain.  These would be interesting research areas to pursue in future studies.
Using telephone interviews, as compared with questionnaires or text messaging, may have had a positive effect on the outcomes obtained, as reported in other studies. [31-33] An attempt to minimize this effect was done by employing anonymous research assistants who collected the data from the university rather than contacting the patients directly from chiropractic offices. We also do not know if these 44 chiropractors who chose to contribute patients for this study are a fair representation of the “typical” Swiss DC. Because the participating DCs seem to be more interested in furthering chiropractic research, they may be a special population. Finally, patients in the subacute category, the least common presentation period in this database, were not included in this study, and there is a need to evaluate this patient group in the future.
In this study, most patients with acute and chronic LBP undergoing chiropractic treatment reported clinically significant improvement. Unlike previously reported studies, patients continued to improve, even after 3 months. As expected, improvement was quicker and more substantial for patients with acute pain. The strongest independent predictor for improvement at 1 and 3 months in both patients with acute pain and patients with chronic pain was whether or not the patient had improved by 1 week. Patients with chronic pain whose LBP was caused by trauma were also more likely to improve at 1 month. Also, unlike previous studies, the presence of radiculopathy in addition to LBP was not a negative predictor for improvement.