Continue Conservative Care or Refer On: The Decision Points in Lumbar Radiculopathy

Most lumbar radiculopathy resolves with conservative management, which is exactly what makes the referral decision hard rather than easy. If every case needed surgery, or every case settled with rest and reassurance, there would be no decision to make. The reality sits in between: a meaningful minority of patients don't improve on the timeline conservative care assumes, and the clinician managing them week to week is usually the one left to notice when the trajectory has stalled. In practice that decision often gets made on gut feel, a bad week for the patient, a difficult conversation, rather than on anything that was tracked from the first appointment.

The problem isn't clinical judgement. It's that judgement without a baseline has nothing to compare against. If you didn't record how much leg pain limited function in week one, "not really improving" in week six is an impression, not a finding. Treating radiculopathy conservatively for six, eight, twelve weeks without a structured re-assessment point is common practice, but it's also why referral decisions get made later than they should, after the patient has lost confidence in the plan rather than because the data crossed a threshold.

Ruling in radiculopathy, and ruling out its mimics

Before any conservative trial has a defined endpoint, the diagnosis needs to be right, and radicular leg pain has more mimics than most clinicians give it credit for. Lumbar spinal stenosis presents with leg pain and neurogenic claudication that can look radicular on a bad day, but its natural history and referral thresholds are different enough that lumping it in with disc-related radiculopathy skews the conservative trial from the outset (Lurie et al., 2016). A patient with genuine stenosis who fails a structured trial of activity modification and loading has a different surgical calculus to a 35-year-old with an acute disc protrusion, and running both on the same clock does neither any favours.

Sacroiliac joint pain is the other common confusion. Posterior thigh and buttock referral from the SIJ is frequently read as radicular, particularly when a formal neurological exam wasn't repeated to confirm a genuine dermatomal pattern, and current guidance on SIJ pain management is clear that it needs its own diagnostic pathway and its own conservative benchmarks, not radiculopathy's (Migliorini et al., 2025). Getting this differentiation right in the first or second appointment matters more than any decision point that comes later, because a conservative trial measured against the wrong diagnosis will never produce a clean answer either way.

What a defensible conservative trial actually contains

A conservative trial that holds up, to the patient, to a referring GP, to whoever reviews the case at six weeks, needs three things: a technique with some evidence behind it, a baseline, and a re-test schedule. Neural mobilisation has a reasonable evidence base for low back and radicular pain, with systematic review data showing benefit for pain and function when added to standard management, which makes it a legitimate part of the trial rather than something added to fill the session (Peacock et al., 2023). What most trials are missing isn't the technique. It's the numbers on either side of it.

At minimum that means capturing the patient's own functional limitations against a baseline, using a tool such as the Patient-Specific Functional Scale, alongside a structured pain measure that captures frequency, intensity and duration rather than a single 0-10 number that moves with the patient's mood on the day. Range of movement, taken with an inclinometer, and strength through the key lower limb muscle groups affected by the suspected nerve root, ankle dorsiflexion, great toe extension, plantarflexion, quadriceps, tested against normative data for the patient's age, sex and weight rather than eyeballed as "4 out of 5", turn a vague impression of progress into a number you can track against an expected rate of change. None of this needs to happen twice a week. It needs to happen at intake and again at a defined re-test point, so week six is being compared to week one rather than to how the clinician remembers week one.

The checkpoints that tell you to refer

Red flags aside, cauda equina symptoms, progressive motor deficit, and unexplained systemic signs are an immediate referral regardless of where a patient sits in a conservative programme, the decision to refer for imaging or a surgical opinion in uncomplicated radiculopathy usually comes down to a handful of measurable checkpoints rather than a single dramatic event:

• No meaningful movement in PSFS or pain score by the defined re-test point. If the change since baseline sits below the expected minimal clinically important difference at six weeks, that's data telling you the current plan isn't working, not a reason to extend it unchanged.
• Strength against normative expectation is flat or declining. A patient whose myotomal strength percentage hasn't moved, or has dropped, between intake and re-test is a different clinical picture from one who's slowly closing the gap, even if both report similar pain scores.
• Progressive neurological signs. Any new or worsening motor deficit between visits moves the conversation from continue and reassess to refer, regardless of how the pain score reads.
• Activity level dropping below recommended thresholds. A patient becoming progressively less active against NHS guideline levels, driven by fear or avoidance rather than genuine incapacity, is a marker that the conservative approach is losing the patient's buy-in, which matters for the referral conversation even before it matters for the biology.

None of these checkpoints require a dramatic escalation to apply. They require being checked on a schedule that was set before the trial started, rather than whenever the patient's frustration forces the conversation.

This decision architecture isn't unique to the low back

The same structure, conservative first, with a defined trial period and objective checkpoints rather than an open-ended wait, shows up across MSK conditions where surgery is an option but not a default. Meta-analysis evidence comparing surgical and conservative management of rotator cuff tears consistently finds outcomes converge over time in appropriately selected patients, which is exactly why the decision hinges on defined checkpoints during the conservative period rather than a blanket rule applied at presentation (Longo et al., 2021) (Ryösä et al., 2017). The ProFHER trial on proximal humerus fractures found broadly comparable outcomes between surgical and non-surgical management, reinforcing that operate-or-don't is rarely decided at the first appointment, it's decided by how the conservative pathway performs against pre-set markers (Handoll et al., 2015).

The lesson for radiculopathy isn't that these conditions behave the same way clinically, they don't. It's that the decision-making architecture, baseline, trial period, objective re-test, defined threshold for escalation, is the same regardless of joint or tissue. Clinicians who build that structure once tend to apply it consistently across a caseload rather than reinventing the referral conversation for every patient.

Building the checkpoint into your caseload

None of this requires new imaging pathways or a change to referral criteria your MDT already agrees on. It requires capturing a baseline on the first visit and setting a re-test date before the trial starts, so the decision to continue or refer is made against numbers rather than against how the last consultation felt. Benchmark PS is built to make that baseline and re-test cheap to capture in a normal appointment, PSFS, pain score, inclinometer range of movement, and handheld dynamometer strength against normative data, tracked longitudinally so the six-week checkpoint is a comparison rather than a guess. If your current radiculopathy pathway doesn't have a re-test date on the calendar from the first appointment, that's the gap worth closing before the next patient walks in.

References

1. Lurie J, Tomkins-Lane C. Management of lumbar spinal stenosis. BMJ (Clinical research ed.). 2016;352:h6234. PubMed
2. Migliorini F, Lucenti L, Bardazzi T et al. Management of sacroiliac joint pain: current concepts. European journal of orthopaedic surgery & traumatology : orthopedie traumatologie. 2025;35(1):208. PubMed
3. Peacock M, Douglas S, Nair P. Neural mobilization in low back and radicular pain: a systematic review. The Journal of manual & manipulative therapy. 2023;31(1):4-12. PubMed
4. Longo UG, Risi Ambrogioni L, Candela V et al. Conservative versus surgical management for patients with rotator cuff tears: a systematic review and META-analysis. BMC musculoskeletal disorders. 2021;22(1):50. PubMed
5. Ryösä A, Laimi K, Äärimaa V et al. Surgery or conservative treatment for rotator cuff tear: a meta-analysis. Disability and rehabilitation. 2017;39(14):1357-1363. PubMed
6. Handoll H, Brealey S, Rangan A et al. The ProFHER (PROximal Fracture of the Humerus: Evaluation by Randomisation) trial - a pragmatic multicentre randomised controlled trial evaluating the clinical effectiveness and cost-effectiveness of surgical compared with non-surgical treatment for proximal fracture of the humerus in adults. Health technology assessment (Winchester, England). 2015;19(24):1-280. PubMed