“Chronic low back pain” is not a diagnosis. It is a description of how long something has hurt. The useful question — the one that decides whether any treatment has a chance of working — is which structure in the back is actually generating the pain.
There are only a handful of realistic candidates, they present differently, and they respond to different things. Most of the disappointment in this field comes from treating the wrong one.
Where the pain is actually coming from
- The facet joints. The small paired joints at the back of the spine that let you extend and rotate. Facet pain is usually worse standing, walking and leaning backwards, better sitting or leaning forward, and often felt as a band across the low back that can spread into the buttock but rarely below the knee.
- The sacroiliac joint. Where the spine meets the pelvis. Pain sits low and to one side, often pointed to with a single finger just inside the back of the pelvis, and is typically worse on standing from a chair, rolling in bed, or single-leg loading like stairs.
- The disc itself — discogenic pain. Deep, central, worse with sitting and forward bending, worse getting out of a car. Distinct from a disc that is pressing on a nerve.
- A nerve root — radiculopathy and sciatica. Pain that travels below the knee in a recognisable line, often with numbness, pins and needles or weakness. This is a nerve problem, not a back problem, even though it starts in the back.
- The deep gluteal muscles, including piriformis. Buttock pain that can mimic sciatica closely, usually worse with prolonged sitting and with the hip in certain rotated positions.
- Spinal stenosis. Narrowing of the space the nerves run through. The signature is pain and heaviness in the legs on walking that eases within a minute or two of sitting or leaning forward — people describe leaning on a shopping trolley.
Two of these are frequently present at once, which is why a scan alone rarely settles it. Nearly every adult spine over forty shows degenerative change on MRI, including in people with no pain at all. The report tells you what your spine looks like. It does not tell you what hurts.
What is offered here — and what is not
This practice performs four image-guided spine injections:
- Facet joint injections — into or around the painful facet joints
- Caudal epidural injections — into the epidural space through the sacral hiatus, for nerve-related pain
- Sacroiliac joint injections
- Piriformis and deep gluteal injections
Platelet-rich plasma is the primary biologic used in all four.
Equally important, here is what is not offered: no intradiscal injections, no radiofrequency nerve ablation, no spinal surgery, no stem cell products sourced from anyone other than you. If your problem needs one of those, you will be told so and referred, not treated with something adjacent to it.
What our own data shows
This practice contributes prospectively to the DataBiologics IMPACT registry, a national orthobiologics outcomes registry. Every treatment is entered, and patients are surveyed over time whether they improve or not.
For lumbosacral and pelvic PRP treatments performed here, from 2021 through 2025:
- 154 cases tracked
- 68% met the responder threshold — defined as at least 50% improvement over their own baseline
- 91% patient satisfaction
- Zero serious treatment-related adverse events — 0 of 154
What this is, stated plainly. It is registry data, not a randomised trial. There is no control group and no blinding. It reflects one practice and one physician. Patients who feel better may be more inclined to complete a survey than those who do not, which can push a number upward. Anyone quoting registry figures without saying that is selling something.
What registry data is genuinely good for is the safety column. Zero serious treatment-related adverse events across 154 spine and pelvis injections is a real number, and it does not depend on who filled in a questionnaire.
What the published trials show — including the ones that go against
Here is where most clinic websites stop being honest, so this section will not.
The randomised trials comparing PRP with corticosteroid in the spine are mixed, and the best-designed recent ones have not favoured PRP.
- A double-blind randomised trial of sacroiliac joint injection published in Pain Medicine found corticosteroid outperformed PRP at every timepoint through six months. Using the same ≥50% threshold, roughly 21% of the PRP group responded at one month against about 80% of the steroid group.
- A randomised, double-blind trial in facet joint syndrome published in Joint Bone Spine concluded that its results did not support the use of PRP for that indication.
- Several prospective, non-blinded studies have reported the opposite, with PRP performing comparably or better than steroid at longer follow-up.
Read together, the pattern is uncomfortable and worth saying out loud: the more rigorous the trial design, the less favourable the result for spine PRP.
So why is our responder rate three times the published trials?
Our registry reports 68% responders by the same ≥50% definition those trials use. The PRP arms of the blinded trials report 21–25%. That gap is too large to wave away, and the honest answer is that some of it is study design — an unblinded registry will always flatter a treatment compared with a blinded trial.
But not all of it, and the likeliest remaining explanation is dose.
PRP is not one thing. It is a category. Two syringes both labelled “PRP” can differ tenfold in the number of platelets they actually deliver, depending on the kit, the spin protocol, the starting blood volume and the patient's own platelet count. Most published trials do not report the dose they delivered, and many use systems that produce a low one.
This is Dr. Buford's published research area. He co-authored a 2026 systematic review and meta-analysis in PM&R finding that PRP outcomes in knee osteoarthritis are associated with the amount of total deliverable platelets. He co-authored a narrative review in Biomedicines on how angiogenesis and tissue repair depend on platelet dosing and bioformulation. He wrote the practitioner's guide to evaluating PRP kits and protocols in the Journal of Orthopaedic Experience & Innovation. And he was part of the expert consensus panel on PRP treatment standards published in Arthroscopy.
A trial that does not measure its dose cannot tell you whether the dose was the problem. That is not a defence of every negative result — it is a specific, testable criticism, made by someone who publishes on it rather than someone who read a summary.
What our own data does show is this. Across 154 lumbosacral and pelvic PRP treatments, 68% of patients reached at least 50% improvement over their own baseline. The PRP arms of the blinded spine trials, using that same threshold, report 21–25%. We believe the difference is dose — and dose is precisely the variable those trials do not report.
Our own cases point the same way. The registry records the platelet dose delivered and follows patients at set intervals, so results can be separated by dose rather than lumped together. In this practice, the higher-dose treatments are the ones associated with better relief and with relief that lasts longer — which is exactly what would be expected if dose is the variable the published trials have been leaving uncontrolled.
And whatever else is argued about relative effectiveness, one thing is not in dispute: PRP produced no serious adverse events in 154 injections and carries no cumulative ceiling. That cannot be said of repeated corticosteroid, and it is the reason we start here rather than finishing here.
Why PRP is offered first — the argument that does not depend on beating steroid
Even taking the trials at their least favourable, there is a sequencing argument:
- Corticosteroid has a cumulative cost. Repeated injections affect blood sugar, bone density and the hypothalamic-pituitary-adrenal axis, and there are local tissue effects. That is why the number you can have is limited. PRP carries no such ceiling.
- Radiofrequency ablation works by destroying a nerve. It can help facet pain, and the medial branch nerve regenerates over roughly six to eighteen months, and then it is done again. It is effective and it is destructive, and patients are rarely told the second part plainly.
- Fusion for discogenic back pain has a famously mixed record. That is not a biologics argument — it is the surgical literature's own conclusion.
- PRP is repeatable and non-destructive. Trying it first forecloses nothing. Steroid, ablation and surgery all remain available afterwards. The reverse is not always true.
When steroid, ablation or surgery is the better answer
Sometimes it plainly is, and saying so is the point of this page:
- Severe acute radicular pain where the priority is getting the inflammation down quickly so the person can function and sleep.
- Progressive neurological deficit — worsening weakness, foot drop, or any bladder or bowel change. That is a surgical evaluation, urgently, not an injection.
- Facet pain that has responded well to diagnostic blocks in someone who wants the longest available relief from a single intervention — ablation is a reasonable choice and it is not offered here.
- Significant instability or deformity, or stenosis with progressive functional loss.
The examination and the imaging decide this, not a preference for one tool.
Common questions
What is actually causing my chronic low back pain?
There are only a handful of realistic sources: the facet joints, the sacroiliac joint, the disc itself, a compressed nerve root, the deep gluteal muscles including piriformis, or spinal stenosis. They present differently — facet pain is usually worse standing and leaning back, sacroiliac pain sits low and to one side, discogenic pain is worse sitting and bending forward, and nerve pain travels below the knee. Two are often present at once, which is why an MRI alone rarely settles it. Nearly every adult spine over forty shows degenerative change on imaging, including in people with no pain at all.
What spine injections do you actually perform?
Four, all image-guided: facet joint injections, caudal epidural injections, sacroiliac joint injections, and piriformis or deep gluteal injections. Platelet-rich plasma is the primary biologic used in all four. What is not offered here: intradiscal injections, radiofrequency nerve ablation, spinal surgery, and any stem cell product sourced from someone other than you. If your problem needs one of those, you will be told and referred rather than treated with something adjacent to it.
What are your own results with PRP in the spine?
This practice contributes prospectively to the DataBiologics IMPACT registry. For lumbosacral and pelvic PRP treatments performed here between 2021 and 2025, across 154 cases, 68 percent of patients met the responder threshold of at least 50 percent improvement over their own baseline, 91 percent reported satisfaction, and there were zero serious treatment-related adverse events. That is registry data, not a randomised trial — no control group and no blinding, from one practice and one physician — and anyone quoting registry figures without saying that is selling something.
How do your results compare with the published PRP trials?
Favourably, and by the same measure. The blinded spine trials define a responder the way we do — at least 50 percent improvement over baseline — and their PRP arms report 21 to 25 percent. Our registry reports 68 percent across 154 lumbosacral and pelvic treatments. Some of that gap is study design, because an unblinded registry will always flatter a treatment relative to a blinded trial. But the trials themselves are mixed, and the variable almost none of them report is the platelet dose actually delivered. That is the difference we think matters, and it is what Dr. Buford publishes on.
Then why is your responder rate so much higher than those trials?
Part of it is study design — an unblinded registry will always flatter a treatment compared with a blinded trial. But the likeliest remaining explanation is dose. PRP is a category, not a product, and two syringes both labelled PRP can differ tenfold in the platelets they actually deliver depending on the kit, the spin protocol and the patient's own platelet count. Most trials do not report the dose they delivered. Dr. Buford publishes on exactly this: he co-authored a 2026 systematic review in PM&R finding PRP outcomes in knee osteoarthritis track with total deliverable platelets, and a review in Biomedicines on platelet dosing and tissue repair. A trial that does not measure its dose cannot tell you whether the dose was the problem.
Why try PRP before a steroid injection or nerve ablation?
Because trying it first forecloses nothing. Corticosteroid has a cumulative cost — blood sugar, bone density, the HPA axis, local tissue effects — which is why the number you can have is limited. Radiofrequency ablation works by destroying the medial branch nerve; it can help, the nerve regrows over roughly six to eighteen months, and then it is done again. PRP is repeatable and non-destructive, and steroid, ablation and surgery all remain available afterwards. The reverse is not always true.
When would you tell me not to have an injection at all?
Progressive neurological deficit — worsening weakness, foot drop, or any change in bladder or bowel function — is an urgent surgical evaluation, not an injection. Severe acute nerve pain sometimes needs the inflammation brought down fast, and steroid does that better. Significant instability or deformity, and stenosis with progressive functional loss, are structural problems. Facet pain that has responded well to diagnostic blocks in someone who wants the longest relief from a single intervention is a reasonable case for ablation, which is not offered here.
