Acute Lower Back
Interesting paper fall little bit short…
Reading Past the Headline: What the PACBACK Trial's Dosing Data Actually Shows About Spinal Manipulation
The PACBACK trial, published in JAMA in December 2025, is already being cited as evidence that spinal manipulation adds nothing for acute low back pain. It's a large, well-run trial, and its headline finding — that clinician-supported self-management modestly outperforms guideline-based medical care — deserves to be taken seriously. But a close read of the trial's own methods and results reveals something the discussion section never addresses: the spinal manipulation arm was dosed at a fraction of what chiropractic clinical practice guidelines recommend, using a protocol the authors themselves describe as not representative of usual practice. That detail changes what the null finding on manipulation can and can't tell us — and it's worth walking through carefully, using the guidelines and dose-response evidence the profession actually relies on.
What PACBACK Set Out to Test
The Spinal Manipulation and Patient Self-Management for Preventing Acute to Chronic Back Pain (PACBACK) trial randomised 1,000 adults with acute or subacute low back pain at elevated risk of chronicity into one of four groups: spinal manipulation therapy alone, clinician-supported biopsychosocial self-management alone, the two combined, or guideline-based medical care. It's a rigorous piece of work by several measures — a 2×2 factorial design, intention-to-treat analysis, 93% completion at one year, and outcomes assessed by staff blinded to group assignment. Its finding that supported self-management produced a small but statistically significant reduction in disability compared with medical care is a genuine, well-supported result. None of what follows is a claim that the trial was poorly conducted overall. The issue is narrower and more specific: whether the spinal manipulation arm, as actually delivered, was capable of answering the question it's now being used to answer.
What the "Spinal Manipulation" Arm Actually Received
The trial's own methods and results sections are unusually specific about this, and the numbers are worth setting out plainly:
- Minimum protocol: "adherent" was defined as attending 2 or more 15- to 20-minute visits across the entire 8-week intervention period.
- Actual mean utilisation: 6.6 visits over 8 weeks — averaging roughly 0.8 visits per week, or one visit approximately every 9 days.
- Adherence: the spinal manipulation arm had the lowest adherence of all four groups at 84%, partly attributed to COVID-19-related visit cancellations affecting nearly 10% of participants in this specific arm.
- Care package: the spinal manipulation arm received manipulation plus soft tissue mobilisation, lumbar neural mobilisation, and moist heat — but not the structured active exercise and self-management component, which was deliberately reserved for the other two arms.
- Delivery: manipulation was provided interchangeably by "the same physical therapists and chiropractors," without separating outcomes by practitioner profession.
What Chiropractic Clinical Practice Guidelines Actually Recommend
Set those numbers next to the dosing standards the chiropractic profession's own evidence-based guidelines describe for acute and subacute low back pain, and the gap is substantial.
| Source | Recommended frequency/dose | Reassessment point |
|---|---|---|
| Globe et al., Clinical Practice Guideline: Chiropractic Care for Low Back Pain (JMPT, 2016) | 2–3 visits per week for 2–4 weeks (up to 12 visits per trial of care) | After an initial round of up to 6 visits, roughly 2 weeks in |
| Bussières et al., Canadian Chiropractic Guideline Initiative (JMPT, 2018) — citing its own supporting trials | Underlying trials used twice-weekly visits in the initial weeks (e.g., 2 sessions in week 1, then weekly; or twice weekly for up to 8 sessions over 4 weeks) | Periodic clinical re-evaluation, with explicit caution against passive-care dependence |
| Haas et al. dose-response randomised trial (Spine Journal, 2014) | Standardised delivery at 3 visits per week for 6 weeks, varying only total session count (0, 6, 12, or 18) | N/A — designed specifically to test dose, not to prescribe a single protocol |
Every guideline and dose-response source above converges on the same picture: a proper evidence-based "trial of care" for acute low back pain concentrates visits at 2–3 times per week over the first two to four weeks, then reassesses. PACBACK is not a guideline and made no dosing recommendation of its own — it's the trial under scrutiny here, not a source of practice standards. What it actually delivered in its spinal manipulation arm was a mean of 6.6 visits over 8 weeks, roughly 0.8 visits per week — about a quarter to a third of the frequency every guideline and dose-response source above describes, spread thinly across 8 weeks rather than concentrated in the early window where manipulation is theoretically expected to do the most.
Why Dose Isn't a Trivial Detail
The Haas et al. trial matters here for a specific reason: it's a dose-response study, built to isolate the effect of visit count while holding frequency constant. It found a statistically significant linear dose-response relationship — participants receiving 12 visits had better short-term pain and disability outcomes than those receiving 6, and the 18-visit group sustained benefits longest at one year. In other words, in the very dose-response literature the chiropractic profession's guidelines draw on, more visits produced measurably better outcomes, not a flat line.
The implication: if visit count moves the outcome in the trial literature specifically designed to test that question, then testing spinal manipulation at roughly half the lowest dose examined in that dose-response curve — and at well under a third of what a chiropractic clinical practice guideline defines as a proper trial of care — is not a neutral methodological choice. It's a design decision that would be expected to bias the result toward the null, independent of whether manipulation has a genuine treatment effect at guideline-recommended dosing.
The Admission That Never Resurfaces
This is the part of the paper most worth reading twice. Buried in the Methods section, the authors write:
"The interventions were designed to isolate the effectiveness of specific treatment approaches and regimens used by physical therapists, chiropractors, and medical clinicians rather than representing usual clinical practice."
That sentence is a direct, explicit acknowledgment that the spinal manipulation protocol does not reflect real-world, guideline-concordant chiropractic care. It appears once, framed as a design rationale in the Methods — and then it does not resurface. The paper's stated Limitations are:
- Patients and clinicians were not blinded, and significant findings were limited to subjective outcomes except for one objective measure.
- The trial population underrepresented Hispanic participants, those without college degrees, and lower-income households.
- The study did not control for time, attention, expectation, or clinician effects.
Dosing intensity relative to guideline-recommended chiropractic care is not listed among them — despite the authors having already told the reader, several pages earlier, that the protocol was not built to represent usual practice. The Discussion goes on to describe spinal manipulation as having a "different mechanism of action vs existing guideline-recommended treatments" and reports no significant benefit over medical care — language that invites a direct comparison against guideline-based care, without disclosing that the manipulation arm itself was delivered at well below guideline-recommended intensity.
A Second, Related Discrepancy
Independent of dosing, the spinal manipulation arm was delivered interchangeably by "the same physical therapists and chiropractors," with no separation of outcomes by practitioner profession or training background. That means the trial cannot, even setting the frequency question aside, support a conclusion specifically about chiropractic management of acute low back pain — at most it speaks to a generic, pooled, under-dosed manual therapy protocol, stripped of its usual accompanying active-care component, delivered by two professions with different manipulation training.
Skill Isn't Interchangeable: What Manipulation Training Actually Requires
That pooling matters more than it might first appear, because delivering a therapeutic, well-controlled spinal manipulation is not a skill acquired by completing a short course or a handful of practice sessions — it is a psychomotor competency the biomechanics literature shows takes sustained, structured training to develop. The World Health Organization's global guidelines on chiropractic education set the baseline for full-time professional training at a minimum of 4,200 student/teacher contact hours over four years, including at least 1,000 hours of supervised clinical training, specifically because producing safe, effective manipulative skill requires that depth of exposure.
The biomechanics research on how this skill is actually acquired backs up why that matters. A controlled study by Triano et al. comparing student cohorts with different levels of prerequisite training found that the group with more extensive foundational preparation performed manipulation significantly closer to an expert reference standard on their very first attempt at a new procedure, concluding that "the duration, extent, and content of prerequisites for learning the dynamic and complex manual skills for spinal manipulation can significantly influence the level of skill attainment even early in the course of training." More recent force-plate studies comparing chiropractic students and interns to licensed, experienced chiropractors have found a similar pattern: experienced practitioners reliably reproduce their own individualised force-time delivery when retested, while students show far greater variability — and, notably, more hours of laboratory practice increased students' confidence without reliably increasing their actual competence at hitting target manipulation forces. In other words, mastery of a controlled, therapeutic thrust is measurably distinct from simply having been shown how to perform one.
What this doesn't claim: this isn't a suggestion that physical therapists are incapable of skilled manipulation — many complete extensive manual therapy training of their own. It's a narrower, evidence-based point: manipulation proficiency is training- and experience-dependent, not a binary someone either has or doesn't after a brief exposure. A trial that pools multiple providers of unreported and unstratified training depth and manipulation-specific caseload, on top of already under-dosing the intervention, has a second, compounding reason to be cautious about generalising its null result to what a specifically and extensively trained chiropractic provider achieves at guideline-recommended frequency.
What This Means — and Doesn't Mean
To be fair to the trial: none of this undermines its central, better-supported finding, that clinician-supported biopsychosocial self-management modestly outperformed guideline-based medical care for disability, with effects that held up over a year. That result was properly dosed (4 to 8 sixty-minute sessions, matching its own intervention logic) and is a reasonable, well-earned conclusion. The problem is specific to how the null finding on spinal manipulation is likely to be used — as evidence that manipulation itself doesn't help with acute low back pain, full stop. That's a broader claim than the trial's own design can support, because the version of manipulation tested here was explicitly, by the authors' own admission, not the version that clinical practice guidelines describe or that dose-response trials have actually evaluated.
The Bottom Line
A trial that tests a treatment at roughly a quarter of its guideline-recommended dose, strips out the active-care component that normally accompanies it, pools two different provider types, and then reports the lowest adherence of any study arm, is not well positioned to conclude that the treatment "doesn't add benefit" in any general sense — it's positioned to conclude that this specific, admittedly non-representative protocol didn't add benefit under these specific conditions. The authors said as much themselves, once, in a single sentence in their methods. That sentence deserved a place in the limitations section and in the framing of the conclusions. Its absence there is the discrepancy worth naming plainly: a study can be well-conducted and still test the wrong dose of the right question, and readers relying on the abstract alone would have no way of knowing that happened here.
Frequently Asked Questions
How often was spinal manipulation actually delivered in the PACBACK trial?
Patients in the spinal manipulation arm attended a mean of 6.6 visits over the 8-week intervention period, roughly 0.8 visits per week — less than once weekly on average, with a protocol minimum of just 2 visits total.
What do chiropractic guidelines recommend for acute low back pain treatment frequency?
The 2016 Clinical Practice Guideline: Chiropractic Care for Low Back Pain recommends 2 to 3 visits per week for 2 to 4 weeks, with reassessment after an initial round of up to 6 visits — roughly three times the frequency PACBACK's spinal manipulation arm actually received.
Does visit frequency actually affect outcomes from spinal manipulation?
A dose-response randomized trial by Haas et al. found a statistically significant linear relationship between visit count and outcomes, with 12 visits outperforming 6 in the short term and 18 visits sustaining benefits longest at one year.
Did the PACBACK authors acknowledge the dosing issue as a limitation?
No. The authors stated in their methods that the interventions were designed to isolate treatment approaches "rather than representing usual clinical practice," but this caveat does not appear among the three limitations formally listed in the discussion section.
Does this mean the PACBACK trial is bad research?
No. The trial was well-designed and its finding that supported self-management modestly outperformed medical care is credible. The issue is specific to the spinal manipulation arm, which was tested at a dose well below what chiropractic guidelines recommend, limiting what its null finding can tell us about manipulation delivered as guidelines intend.
Can this trial be used to argue chiropractic care shouldn't be offered for acute low back pain?
Not on the manipulation-specific finding. The trial tested a pooled, under-dosed manipulation protocol delivered by both physical therapists and chiropractors, without the active-care component guidelines typically pair with it, at roughly a quarter of guideline-recommended frequency — conditions the study's own authors describe as unrepresentative of usual practice.
Does it matter which type of provider delivers the manipulation?
Yes, based on the skill-acquisition literature. Manipulation proficiency is training- and experience-dependent: studies show more extensive prerequisite training produces performance closer to an expert standard, and that practice hours alone can boost a trainee's confidence without improving their actual force control. PACBACK pooled physical therapists and chiropractors without reporting or stratifying by training depth or manipulation-specific experience.
- Bronfort G, Meier EN, Leininger B, et al. Spinal manipulation and clinician-supported biopsychosocial self-management for acute back pain: the PACBACK randomized clinical trial. JAMA. 2025;335(6):497-510.
- Globe G, Farabaugh RJ, Hawk C, et al. Clinical practice guideline: chiropractic care for low back pain. Journal of Manipulative and Physiological Therapeutics. 2016;39(1):1-22.
- Bussières AE, Stewart G, Al-Zoubi F, et al. Spinal manipulative therapy and other conservative treatments for low back pain: a guideline from the Canadian Chiropractic Guideline Initiative. Journal of Manipulative and Physiological Therapeutics. 2018;41(4):265-293.
- Haas M, Vavrek D, Peterson D, Polissar N, Neradilek MB. Dose-response and efficacy of spinal manipulation for care of chronic low back pain: a randomized controlled trial. The Spine Journal. 2014;14(7):1106-1116.
- Haas M, Groupp E, Aickin M, et al. Dose response for chiropractic care of chronic low back pain. The Spine Journal (related dose-response series).
- Qaseem A, Wilt TJ, McLean RM, Forciea MA. Noninvasive treatments for acute, subacute, and chronic low back pain: a clinical practice guideline from the American College of Physicians. Annals of Internal Medicine. 2017;166(7):514-530.
- Trial cited within Bussières et al. 2018 (Fritz et al.): early physical therapy including manipulation delivered at approximately 2 sessions in week 1, then weekly.
- Trial cited within Bussières et al. 2018 (Brennan et al.): manipulation delivered twice weekly for up to 8 sessions over 4 weeks.
- World Health Organization. WHO guidelines on basic training and safety in chiropractic. Geneva: WHO; 2005.
- Triano JJ, Rogers CM, Combs S, Potts D, Sorrels K. Procedural skills in spinal manipulation: do prerequisites matter? The Spine Journal. 2004;4(5):557-563.
- Cross-sectional study. Comparison of force-time characteristics of prone cervicothoracic spinal manipulative therapy between chiropractic interns and chiropractors. Chiropractic & Manual Therapies. 2025.
- Secondary analysis. Chiropractic students' characteristics influencing confidence and competence in modulating spinal manipulation force-time characteristics of specific target forces. Chiropractic & Manual Therapies. 2025.
