Back pain is usually treated with the expectation that, with time and rehabilitation, things will improve.
But what happens when it doesn’t?
A recent study following people with chronic low back pain over two years found that most remained on relatively stable pain trajectories, while only a smaller group gradually improved (Rabiei et al., 2026).
Although we’re taught 90% of pain goes away in 6 weeks, this isn’t supported in the literature. Here’s the natural history of back pain based on the literature:
About 20% recover in full
55% have ongoing pain of low intensity and get on with life
15% continue to suffer moderate pain
10% remain significantly disabled
(Artus et al, 2014, Klenerman et al, Spine 1995, Von Korff et al, Spine, 1996)
While most back pain presentations can initially be managed non-specifically and improves on its own accord, there are a group of patients who don’t get better as expected, or whose pain becomes highly intrusive and doesn’t go away.
This is when we need to take “non-specific” back pain to “specific” by identifying the most likely pain generator.
There are five main potential pain generators that can be investigated through diagnostic blocks: the facet joints, sacroiliac joints, hips, discs and nerves. Muscle weakness and wasting can also play an important role in perpetuating pain over time.
Because their referral patterns often overlap, the location of pain can point us in a direction, but it does not necessarily tell us which structure is responsible. That’s why I look at the broader pain pattern, along with the history, examination and imaging, to work out which diagnosis fits best.
How Pain Patterns Help Narrow Down the Cause
Recognising a pain pattern starts with these questions: Where is the pain? Where does it travel? What triggered it? What makes it better or worse? Are there associated symptoms? To add more structure to this process, I use the 5 Minute Low Back +/- Leg Pain Algorithm to work through these questions systematically.
For example, if leg pain is worse than back pain, particularly with a specific distribution and burning, pins and needles, numbness, weakness or altered reflexes, I start thinking more about nerve root involvement. But aching pain travelling into the leg does not automatically mean sciatica. The facet joints, sacroiliac joints, hips, discs and nerves can all refer aching pain into the leg.
The key is to look at which structure best matches the overall pattern
Facet pain, in comparison, is generally an ache above the belt line, commonly more to one side, and may be worse with bending backwards. But can radiate down the leg. True sciatic pain can be electric and follow a recognisable nerve-root distribution.
These are clues. I use them with the examination to form a working diagnosis, then look at whether the imaging supports it.
When Imaging Doesn’t Explain the Back Pain
Structural changes can appear on imaging in people without pain. Finding an abnormality on an MRI therefore does not automatically identify the pain generator.
For unexplained back pain, or when the suspected source still needs clarification, targeted diagnostic procedures can sometimes help test the hypothesis.
If I suspect facet joint pain, for example, a medial branch block temporarily blocks the nerve supply to the joint. A positive response makes facet pain more likely; a second positive block increases diagnostic certainty.
This reflects the broader way I work through lower back pain: start with the structure suggested by the history, examination and imaging, test it where appropriate, then reassess.
If the response does not support the hypothesis, I reconsider the other potential pain generators and move onto the next most likely of the top five.
From Back Pain Diagnosis to Treatment
Once the likely pain generator is clearer, treatment can be more specifically targeted to the underlying source of pain. Where diagnostic blocks support facet joint pain, radiofrequency neurotomy may be considered. Under X-ray guidance, a radiofrequency probe is positioned beside the relevant medial branch nerves and uses heat to interrupt their ability to transmit pain signals.
Radicular pain (true sciatica) follows a different pathway. Where the clinical picture supports nerve-root involvement, epidural, then pulsed radiofrequency are interventional options I may use. Pulse radiofrequency, unlike thermal radiofrequency neurotomy, applies intermittent electrical pulses without heating the nerve in the same way.
The procedure follows the clinical reasoning, rather than replacing it.
With my background in Sports and Exercise Medicine, exercise and rehabilitation are also an important part of how I approach lower back pain. My goal is always to help reduce pain and get people back to movement: returning to exercise, playing sport, spending time with their kids or simply getting back to the things they enjoy.
If someone is getting better, there may be no reason to keep looking for a more specific answer. But when the pain is highly intrusive or someone simply isn’t getting better, I want to understand why.
That answer can be found by understanding the pain pattern, identifying the likely pain generator, testing that hypothesis where appropriate, and using what I learn to guide the next decision.
If you want to see that reasoning laid out visually, explore my 5 Minute Low Back +/- Leg Pain Algorithm and Low Back Pain Algorithm.

