Sciatica: the questions you ask every time

Clinical Rounds, the GPAtlas podcast. Musculoskeletal (MSK). 19 minutes. Published 31 August 2026. Free to listen.

Clinical source: our own Ocean article "Sciatica", last reviewed 8 June 2026.

Transcript

Sarah: Hello, and welcome to GPAtlas Clinical Rounds. I'm Sarah, and this is the podcast where we take one thing you'll genuinely see this week and work out what to do with it.

Ben: And I'm Ben. Today, sciatica. Which is common, usually gets better, and contains one thing you cannot afford to miss.

Sarah: Dev is forty-four, he does deliveries, and he's had left leg pain for two weeks. It goes from his buttock down the back of his leg to his calf, it's worse than his back pain, and it's agony when he coughs. He's come in wanting a scan and something stronger than ibuprofen, and he's already looked it up and decided it's a slipped disc.

Ben: Three things. The screen you do at every single contact, and why the version in your head is probably the late one. Why the scan he wants would most likely make things worse. And the drugs that feel right for nerve pain and are specifically not recommended.

Sarah: So first, is it even sciatica. Our article's description is clean: suspect it when leg pain dominates over back pain, follows a nerve-root distribution, and is aggravated by coughing, sneezing or sitting. Dev has all three. Leg worse than back is the signature, and it's the one that separates this from ordinary low back pain.

Ben: And it frames the whole appointment in a way I find useful. Our article says the two essential primary-care tasks are to recognise the radicular pattern and, at every contact, to exclude the serious mimics, above all cauda equina syndrome. Two jobs. Not diagnose precisely, not image, not cure. Recognise the pattern, exclude the disaster.

Sarah: And the good news you can give him honestly, which he won't have found online: most cases settle within six to twelve weeks with activity and simple measures. That's the single most therapeutic sentence in the consultation, and it's true.

Ben: Leg worse than back. And two jobs, not four.

Sarah: Now the important part, and I want to slow down here because this is the bit that changed how I consult. Cauda equina. Everybody knows the red flags. Saddle anaesthesia, retention, incontinence, bilateral leg symptoms, loss of anal tone. And you ask about those and he says no, and you relax.

Ben: Except our article says this, and it's the line I'd put on the wall: early cauda equina is easy to miss because the textbook picture, established saddle anaesthesia with painless retention and overflow, is a late finding. Late. Everything most of us screen for is the late presentation. By the time he can't feel himself sitting on a saddle, you have missed the window you were screening for.

Sarah: So what are the early ones? Our article lists them: difficulty starting to pass urine, a reduced sense of bladder filling, and loss of the normal sensation of urinary flow. Those are the softer warnings. And none of them is what a patient would think to mention. Nobody rings the surgery to say they can't feel their urine flowing the way they used to. You have to ask.

Ben: And the action attached is unambiguous. Early bladder symptoms, so difficulty initiating or loss of awareness of filling or of the stream, may be early cauda equina before painless retention develops, and that is a same-day assessment. The full picture, bilateral leg pain or weakness, saddle anaesthesia, retention or incontinence, faecal incontinence or loss of anal tone, is emergency same-day admission for urgent MRI. That's 999 or the emergency department.

Sarah: And our article tells you what to do with your consultation, not just your brain. Ask about these directly in any bilateral or severe sciatica. Document perineal sensation and post-void status. And safety-net explicitly for same-day assessment, not wait and see. Those last four words are there because wait and see is exactly what happens otherwise.

Ben: It even gives you the words for the patient, which I think is worth reading out as written. Please come back urgently if you develop numbness around the back passage or genitals, trouble passing urine, or new weakness in both legs. That's the sentence. Say it, and write that you said it.

Sarah: The textbook picture is the late one. Ask about the flow.

Ben: One other red flag set, because sciatica-shaped pain isn't always a disc. Spinal malignancy or cord compression: known cancer, unexplained weight loss, age over fifty, thoracic pain, or night or rest pain. Any of those and you're on a suspected cancer pathway, not a back pain pathway. Thoracic pain in particular doesn't belong to sciatica at all.

Sarah: And progressive weakness. A worsening foot drop or rapidly evolving weakness or numbness is an urgent spinal or neurosurgical referral. So the thing to ask at review isn't just is the pain better, it's has anything got weaker. Pain that improves while power quietly worsens is a bad combination that patients often report as improvement.

Ben: Thoracic pain is not sciatica. And ask what has got weaker.

Sarah: So, examination. And it's quicker than people think. Gait and posture, including heel- walking, which tests L4 and L5 dorsiflexion, and toe-walking, which tests S1 plantarflexion. Look for a list or a foot drop. That's most of your neurology, done watching him walk from the door to the chair.

Ben: And the straight leg raise, which has more precision in it than it usually gets. It counts when it reproduces radicular pain down the leg, not just back pain or hamstring tightness, and between roughly thirty and seventy degrees of hip flexion. Outside that range, and without leg pain, it isn't a positive test.

Sarah: With the caveat that saves you from over-reading it: a normal examination, including a negative straight leg raise, does not exclude sciatica. So it isn't a rule-out. It's one more piece of a picture that's mostly built from the history.

Ben: Thirty to seventy degrees, and it has to be leg pain.

Sarah: Now the scan, which is what he came for. Our article is direct: sciatica is a clinical diagnosis, and you do not offer imaging in primary care, certainly not in the first four to six weeks, because it rarely changes early management.

Ben: And there's a better reason than rationing, which is the one to actually say to him. MRI commonly shows disc bulges and degenerative change in people with no symptoms at all, so a scan in isolation can mislead. You image only when the result will change management, and you always read it alongside the clinical picture.

Sarah: Which gives you something honest to say instead of no. If I scan you today, there's a good chance I'll find a bulge, and there's a good chance that bulge is also in half the people your age who have no pain whatsoever. And then we'd both be looking at a picture that doesn't tell us what to do. That's a different conversation to computer says no.

Ben: And there's a genuinely reassuring fact in our article that patients almost never hear. Most disc herniations shrink and resolve over months, and the larger herniations often recover best. The big one on the scan is not the worst news. That's counter-intuitive enough to be worth telling him.

Sarah: The bulge is in the pain-free people too. And the big ones do best.

Ben: Analgesia, and here's where habit and evidence part company. What is recommended: an NSAID if not contraindicated, ibuprofen or naproxen, lowest effective dose for the shortest period, with a proton pump inhibitor for gastroprotection. Codeine comes second line.

Sarah: And what is not recommended, which is the list most of us have prescribed from. Gabapentinoids, so gabapentin and pregabalin. Other antiepileptics. Oral corticosteroids. And benzodiazepines. Our article's words are no overall benefit and evidence of harm. Not neutral. Harm.

Ben: And it names why we do it, which is fair, because the logic is seductive. Because sciatica is neuropathic, it is tempting to prescribe gabapentin, pregabalin or amitriptyline, and some local pathways and habits still do. The reasoning is sound and the evidence doesn't follow it. Nerve pain does not automatically mean nerve pain drug.

Sarah: And a short list of things not to bother with, so you can answer when he asks: no traction, acupuncture, TENS, ultrasound, belts or corsets, or foot orthotics. Limited or no evidence of benefit. That saves him money as much as anything.

Sarah: Neuropathic pain does not mean a neuropathic drug here.

Ben: What does help is dull and it works. Reassure and explain the favourable natural history, most settle within six to twelve weeks. Encourage staying active, and staying at or returning to work. And advise against prolonged bed rest.

Sarah: And for Dev that work conversation is the consultation, really. He does deliveries. Telling him to stay active is easy for me to say and hard for him to do, and if I don't engage with what his actual day looks like, he'll either ignore me or he'll sign off for six weeks and deconditioning will do more damage than the disc.

Ben: Then the referral threshold. Persistent radicular pain beyond about six weeks despite conservative care goes for a spinal, musculoskeletal or neurosurgical opinion, for injection or decompression. Six weeks. And a point our article makes that I think is worth flagging: a surgical referral should not be withheld because of weight, smoking status or psychological distress.

Sarah: Six weeks. And do not gatekeep the referral on his weight.

Sarah: There's one more thing I want to say about Dev, and it's about the shape of the appointment rather than the medicine. He came in with a diagnosis, a request for a scan, and a request for stronger painkillers. And on the face of it I'm refusing all three: I'm not scanning him, I'm not giving him gabapentin, and I'm telling him to keep working.

Sarah: So if I don't handle that well, he leaves thinking I've done nothing. And the way I've learned to avoid that is to give him the things he can't get from the internet. A number, six to twelve weeks. A reason the scan would mislead. A specific list of symptoms that mean come straight back. And a plan for what happens at six weeks if it hasn't shifted.

Sarah: That's four concrete things, none of which is a prescription, and it turns a refusal into a plan. Whereas no scan, no strong painkillers, keep working is the same clinical decision and it feels to him like being dismissed.

Ben: And it's worth noticing that the safety-net is the part he'll actually act on, so it has to be specific. Numbness around the back passage or genitals, trouble passing urine, new weakness in both legs. Not any problems, come back. Those exact symptoms, and urgently.

Sarah: Four concrete things turn a refusal into a plan.

Ben: Worth restating the whole thing as a sequence, because it's short and it holds under pressure. Recognise the pattern: leg worse than back, radicular distribution, worse on coughing. Screen for cauda equina, and screen for it every time, including the urinary questions that come before the classic signs. Look for the cancer flags: known cancer, weight loss, over fifty, thoracic pain, night pain.

Ben: Then examine briefly, mostly by watching him walk, with heel and toe walking and a straight leg raise that only counts between thirty and seventy degrees with genuine leg pain. Don't image in the first four to six weeks. Treat with an NSAID and a proton pump inhibitor, not with a gabapentinoid. Keep him active and at work. And refer at six weeks if the radicular pain persists.

Sarah: And the one that runs through all of it: none of your reassuring findings are actually reassuring. A normal examination doesn't exclude sciatica. A negative straight leg raise doesn't. Normal perineal sensation doesn't exclude early cauda equina. Chest wall tenderness, in the other episode, doesn't exclude a heart attack. There's a pattern to what goes wrong in medicine, and a lot of it is treating the absence of a late sign as evidence.

Ben: None of the reassuring findings are reassuring. That is the pattern.

Sarah: And the last thing I'd say about Dev is about what happens at six weeks, because that review is the one most likely to be missed. He'll either be better, in which case nothing needs to happen, or he won't, and that's the referral point. If nobody books it, the default is that he comes back at four months having quietly deteriorated, and by then the conversation is much harder.

Sarah: So I'd book it rather than safety-net it. Come back in six weeks if it hasn't settled is an instruction that relies on him deciding whether six weeks has passed and whether not- quite-better counts. A booked appointment that he can cancel if he's fine is a much more reliable way of hitting the referral threshold our article sets.

Ben: Book the six weeks. Do not leave it to him to judge.

Ben: Exam corner. A fifty-two year old man has had right-sided sciatica for three weeks, improving slowly. At review he mentions that for the past two days he has had to wait longer than usual to start passing urine, and he isn't always sure when his bladder is full. His legs feel the same as before and perineal sensation is normal. What is the most appropriate action? A, reassure and review in three weeks. B, arrange a routine MRI. C, start gabapentin for neuropathic pain. Or D, arrange same-day assessment.

Ben: It's D. Difficulty initiating micturition and loss of awareness of bladder filling are the early cauda equina symptoms, and our article says they may precede painless retention, so they warrant same-day assessment. The trap is that his legs and his perineal sensation are normal, which is exactly what early cauda equina looks like, because the textbook findings are late ones. A is the error the article specifically warns against with the phrase not wait and see. B is far too slow and routine imaging is not the pathway here. And C is on the not-recommended list anyway.

Sarah: So. Three things for Monday. One. Ask about the flow, not just the numbness. Difficulty starting, and losing the sense of the bladder filling, come before everything you were taught to look for. Two. Don't scan in the first four to six weeks, and explain why in terms of what a scan would find in a pain-free person. And three. If you're reaching for gabapentin because it's nerve pain, our own article says no overall benefit and evidence of harm.

Sarah: One thing to reflect on, if you're logging this. Think about your last sciatica consultation. Did you ask about starting to pass urine, and did you write down that you asked? The full transcript and the references are on the episode page. Ben and I are synthetic voices. The medicine isn't. See you next week.

More from Clinical Rounds

Every episode of Clinical Rounds, free to listen with a full transcript.