Painkillers and clear thinking

Prepared 4 September 2026 by Kya for her partner and the patient. It answers one question: which painkillers are least likely to affect the patient's thinking and communication, and what plan follows from that. Figures come from Cochrane reviews, NICE guidance and the trials named in the sources. Where something is judgement rather than evidence, it says so. This is a list of things to ask the doctors for, not instructions to self-medicate. The full research with about 40 sources is in research thread 08.

The short answer

1. Uncontrolled pain and broken sleep are already costing the patient her clarity. Chronic pain measurably impairs attention, memory and word-finding, and one bad night raises pain sensitivity the next day. "No drugs" is not the safe option. It is the one she has now.
2. The most cognitively damaging thing in her current regimen is hourly as-needed Oramorph. Each immediate-release dose impairs memory for hours (Kamboj 2005), and the peaks and troughs never let tolerance develop. A small, steady, slow-release background dose is kinder to her thinking than the same total given in bursts, and over a year of steady dosing patients' cognition did not worsen (Tassain 2003).
3. Carbamazepine is the drug most likely to produce what her partner fears. Two thirds of people get a side effect, about a quarter report memory and concentration problems, women are affected more, and it interferes with dexamethasone and later treatments. Ask for oxcarbazepine, duloxetine or nortriptyline instead.
4. Paracetamol, anti-inflammatories, lidocaine plaster and nerve blocks cost nothing cognitively. Use them fully before anything that acts on the brain.
5. One new central drug at a time, 48 to 72 hours apart, with a written stop rule for each. That is how you tell a drug effect from a pain effect and reverse it the same day.

The drugs, ranked by effect on thinking

Band A: essentially none

• Paracetamol, regular.
• Lidocaine 5 percent plaster to the cheek and temple; lidocaine nasal spray for flares.
• Sphenopalatine ganglion block (bedside, through the nose) and occipital nerve block.
• Anti-inflammatories: diclofenac (started today), naproxen, ibuprofen, celecoxib. Their risks are to the kidney, stomach and platelets, all monitorable. At 51 kg with low overnight blood pressure she needs a creatinine check now and again in a few days, plus a stomach-protecting tablet. Celecoxib does not affect platelets, so it is the choice for the days before a biopsy rather than stopping anti-inflammatories entirely.

Band B: small, mostly in the first week or two, mostly avoidable by going slowly

• Duloxetine. Least sedating of the nerve-pain drugs; NICE first line. Watch nausea, appetite and any emotional flattening.
• Nortriptyline or amitriptyline at night, low dose. Morning grogginess settles in a week or two and it helps sleep.
• Low steady background opioid (slow-release morphine or oxycodone, 10 to 15 mg twice daily) with a laxative. Some drowsiness in the first days, then tolerance.

Band C: real, dose-related, improves over weeks but does not vanish

• Gabapentin and pregabalin. Sleepiness in 11 to 15 percent and dizziness in 13 to 22 percent at working doses (Cochrane), against 3 to 5 percent on placebo. They add to opioid sedation.
• Oxcarbazepine. Fewer side effects than carbamazepine, but low sodium is common and is a hidden cause of confusion. Sodium must be checked.

Band D: the ones most likely to change her

• Carbamazepine (see above).
• Hourly as-needed Oramorph, because of the dosing shape, not the drug.
• Ketamine: dissociation and confusion are its signature. Last resort.
• Dexamethasone: about a third of people have a psychological reaction, mostly sleeplessness and agitation. It is being held for the biopsy anyway, which is correct.

Ranking within a band is judgement. The bands themselves rest on the trial figures.

The plan to ask for

Step 0, today: the foundations

• Regular paracetamol (pharmacist to confirm the dose at 51 kg).
• Regular anti-inflammatory with a stomach protector, and a creatinine check from the 3 September bloods, repeated in 3 to 5 days. Ask about naproxen for the longer term and celecoxib for the pre-biopsy window.
• Lidocaine plaster on the right face, 12 hours on and 12 off; lidocaine nasal spray for flares.
• Laxative and anti-sickness prescribed alongside any opioid.
• A pain score at every set of obs, and "worst pain in the last 24 hours" at every round.

Step 1, today or tomorrow: reshape the opioid

Ask: "Her pain is continuous. Could she have a low regular slow-release opioid rather than hourly as-needed Oramorph, with the as-needed dose kept for genuine breakthrough?"

• Slow-release morphine 10 mg twice daily to start (oxycodone if kidney function is reduced), reviewed daily. Rescue Oramorph counted.
• Rule: more than 3 rescue doses a day, raise the background by about a third. Drowsy or confused, reduce. No benefit at a moderate dose, this is not an opioid-responsive pain and the answer is a nerve-pain drug, not more opioid.
• If the neck jolts get worse as the opioid becomes regular, say so. Opioid myoclonus is a reason to switch or reduce.

Step 2, within 48 hours: one nerve-pain drug, chosen for cognition

First ask the team to characterise the pain: "Is there a stabbing, electric, touch- or chewing-triggered component, or is it all continuous ache with flares?" That decides the drug.

• Continuous ache (what her partner describes): duloxetine 30 mg each morning, to 60 mg after a week. Watch nausea, appetite and weight, sodium at two weeks, and whether she feels like herself. Stop if appetite loss persists past two weeks, if any emotional blunting is noticed, or if no benefit at 60 mg by four weeks.
• Sleep the bigger problem: nortriptyline (or amitriptyline) 10 mg two hours before bed, up by 10 mg every 3 to 5 days to 25 to 50 mg. Stop back a step if morning fog lasts past a week. Not with duloxetine.
• Clear trigger-evoked stabbing component: oxcarbazepine rather than carbamazepine, 150 mg twice daily, up slowly, with sodium checked before starting, at one to two weeks and after each increase. Any rash: stop at once. Sodium under 130: stop or reduce.

What the treating team chose, 5 September. The Southampton neurosurgery registrar took a full history, found no single nerve that accounts for all of the pain, and recommended gabapentin rather than a nerve block. That reasoning is sound: a clival and cavernous sinus mass irritates several nerves, the dura and the bone at once, and gabapentin is NICE first-line for that kind of pain. The family accepted it with the conditions from this document: start at 100 to 300 mg at night, step up every few days rather than daily, keep the morphine dose modest while it builds, and an agreed stop rule, which is drowsiness or fog lasting more than a week at a steady dose, at which point it is swapped for duloxetine or nortriptyline rather than pushed. It takes days to work, so regular morphine and ibuprofen still matter in the meantime.

What Southampton provided on 5 September, and the interim. The department prescribed Oramorph, gabapentin, diclofenac suppositories, omeprazole and a laxative to take home, but not slow-release morphine. Until a GP or the pain team converts to a slow-release base, the family can approximate one by giving the Oramorph by the clock at fixed intervals within the prescribed frequency, whether or not she is asking for it, and logging every dose; that keeps her ahead of the pain and gives the next prescriber an exact daily total to convert. An anti-sickness tablet should sit alongside the morphine as the laxative does.

Gabapentin: the full brief is a separate document, "the patient: gabapentin, before you decide to start it" (5 September), with the evidence figures, the slow schedule, the side effects with rates, the morphine interaction, the diary and the hold, drop and stop rules, and ten questions for the prescriber. The summary here stands.

Gabapentin in detail, for the family. It damps overactive nerve signalling by binding a calcium channel subunit in nerves and spinal cord; it is neither an opioid nor an anti-inflammatory, so it is added to those, not swapped for them. NICE first-line for neuropathic pain; in the Cochrane review about one person in six to seven gets a halving of pain at a full dose and more get a lesser benefit; it takes one to two weeks at a working dose to judge. Side effects: sleepiness in roughly one in eight and dizziness in one in five at working doses, worst in the first weeks and at each step up; some weight gain and ankle swelling over time; with an opioid it adds to sedation and can slow breathing (the MHRA warning), so the opioid stays modest while it climbs. It is a controlled drug in the UK because of misuse elsewhere. The trade for the patient: the drug on this list most likely to fog her and also the most likely to help a diffuse, multi-nerve pain. The way to have both: the slow climb (300 mg at night for three nights, then twice a day, then three times a day, stepping up only if she is clear-headed) and the stop rule (fog or drowsiness lasting more than a week at a steady dose means a swap to duloxetine or nortriptyline, not a higher dose). Never stopped suddenly after weeks of use; tapered over a week or more.

Step 3, only if step 2 fails or is not tolerated: pregabalin or gabapentin

Started low and slowly, slower than the label because she is on an opioid (MHRA warning about breathing). If fog or drowsiness persist two weeks at a steady dose, drop one step. If no benefit at a fair dose after two weeks, taper off.

Step 4, any time a flare is uncontrolled: the acute pain team's tools

• Sphenopalatine ganglion block: ask on day one, repeatable, no cognitive cost.
• Occipital nerve block for the neck component.
• Intravenous lidocaine as a monitored rescue.
• Patient-controlled analgesia during a bad flare: it removes the waiting and the pacing, and the total dose is usually lower.
• Ketamine only if everything else has failed, with the family told what it does to perception.

Non-drug measures that protect her thinking

• Sleep. Ask the consultant to write in the plan that overnight obs are spaced if her early-warning score allows, night doses timed so she is not woken, lights and noise down. This is the cheapest cognitive protection there is.
• Heat to the neck and jaw for the spasm. Safe, immediate, hers to control.
• Neck position. Head supported in neutral, one firm pillow, left side-lying with a pillow between the knees, not flat on the back with the head turned. Judgement, not citation.
• Physiotherapy for the guarding spasm. Ask.
• Eye patch or Fresnel prism for the double vision, from the orthoptist. Diplopia is exhausting and a headache source in itself.
• A two-line daily diary kept by her partner. Time, pain score, what was given, hours slept, and what the patient was like: clear, foggy, flat, herself. It is how the team tells drug effect from pain effect.
• Dietitian, for the weight loss and to protect the kidneys under an anti-inflammatory.

Who to ask for, and the words to use

• Acute pain team, today. The Faculty of Pain Medicine standard is assessment of complex inpatient pain within 24 hours. Ask: "What is the regular plan, which nerve-pain drug, what is the stop rule for each, can she have a sphenopalatine block, and who reviews her tomorrow?"
• Supportive (palliative) care team, now. Say: "We understand this is for symptom control, not end of life." They manage exactly this combination every day, and they write the discharge plan and hold it in the community.
• Pharmacist for a medicines review: interactions, kidney dosing, the paracetamol dose at her weight.
• The named consultant to write the plan and the stop rules into the notes, so night staff can act instead of declining a review.

One paragraph for the ward: "the patient's pain is continuous, right trigeminal territory, with an inflammatory component that responded to anti-inflammatories for a year and a nerve component that codeine and as-needed morphine do not touch. Her family's overriding concern is preserving her thinking and communication. Please could she have regular paracetamol and an anti-inflammatory with stomach protection and a creatinine check; a low regular slow-release opioid with rescue doses instead of hourly as-needed Oramorph; one nerve-pain drug started now, chosen for lowest sedation; lidocaine plaster and nasal spray; a sphenopalatine ganglion block this week; acute pain team and supportive care referrals today; overnight obs spaced to protect sleep; and a written stop rule for each drug so that any drowsiness or confusion is acted on the same day."

Bone: what the pain drugs do and do not do

Asked by her partner on 5 September because the tumour is destroying bone. Timescale is everything: none of the effects below matters over days, and nothing protects the clivus itself except treating the tumour.

• Dexamethasone slows bone formation and speeds resorption; long courses cause osteoporosis. A short course for swelling does not measurably change what the tumour does locally. The stronger reasons to hold it now are that it can spoil a lymphoma biopsy and does not shrink a chordoma. If steroids are ever needed for more than a couple of weeks, ask for calcium and vitamin D cover and a bone-protection plan.
• Diclofenac and other anti-inflammatories: no effect on bone density; they can slow bone healing after surgery, so surgeons often pause them for a few weeks post-operatively.
• Opioids: months of use lower sex hormones, which thins bone, and cause falls. Short use, nothing.
• Carbamazepine and phenytoin: speed up vitamin D breakdown and thin bone over years; another reason to prefer oxcarbazepine or duloxetine.
• Duloxetine, amitriptyline, nortriptyline: modest bone loss and fracture signal with years of use.
• Gabapentin, pregabalin, diazepam: falls rather than bone.
• Omeprazole: years of use reduce calcium absorption slightly.
• Paracetamol: none.

What protects her skeleton: vitamin D and calcium to the normal range after testing; protein and weight; walking and standing; replacing any pituitary hormones that the 3 September profile shows to be low. Bone drugs (bisphosphonates, denosumab) have only isolated case reports in chordoma, slow bone healing and can cause jaw bone death, so they are not for now; they belong in a steroid bone-protection conversation if months of steroids are ever needed.

Red flags to watch for, by drug

Tell the nurse in charge the same hour. If unhappy with the response, the registrar, then the consultant.

• Any opioid: hard to wake, breathing slow (under 10 to 12 a minute) or shallow, pinpoint pupils, confusion or hallucinations, new or worse jerking, no bowel movement for three days.
• Anti-inflammatories: black or tarry stool, vomiting blood, stomach pain, passing much less urine, ankle swelling, wheeze. Not drinking for a day while on one is itself a flag.
• Oxcarbazepine, carbamazepine, lamotrigine: any rash, mouth ulcers, fever or sore throat (stop and tell the doctor at once); headache with nausea, confusion, unsteadiness or new drowsiness (low sodium, blood test that day); yellow skin or eyes; unusual bruising.
• Gabapentin, pregabalin: heavy drowsiness, slow breathing especially with an opioid, falls, sudden mood change.
• Nortriptyline, amitriptyline: fast or irregular heartbeat, fainting on standing, cannot pass urine, severe constipation, marked confusion.
• Duloxetine: agitation, sweating, fever, tremor, twitching, racing heart (serotonin syndrome, especially if tramadol, metoclopramide or ondansetron is added); confusion with headache and nausea (low sodium); black stool with the anti-inflammatory; a sudden emotional change. Do not stop it abruptly.
• Dexamethasone, if started after biopsy: sleeplessness, agitation, euphoria or low mood, thirst and passing more urine, thigh weakness. Morning dosing helps.
• Overriding, whatever the drug: any change in sight or eye movement, new facial numbness or weakness, sudden severe headache with vomiting or drowsiness, fever with a stiff neck, new limb weakness.

Key sources

• NICE CG173 Neuropathic pain in adults: pharmacological management.
• NICE CG140 Palliative care for adults: strong opioids for pain relief.
• NICE CKS Trigeminal neuralgia; Bendtsen et al 2019, European Academy of Neurology guideline on trigeminal neuralgia.
• Cochrane reviews: gabapentin (Wiffen 2017), pregabalin (Derry 2019), carbamazepine (Wiffen 2014), amitriptyline (Moore 2015), duloxetine (Lunn 2014).
• Kamboj et al 2005, Pain: immediate-release opioid doses and memory. Tassain et al 2003, Pain: sustained-release morphine and cognition over 12 months.
• Moriarty et al 2011, Progress in Neurobiology: chronic pain and cognition. Krause et al 2019, Journal of Neuroscience; Finan et al 2013, Journal of Pain: sleep and pain.
• Leese et al 2000: celecoxib and platelet function.
• Faculty of Pain Medicine, Core Standards for Pain Management Services in the UK. MHRA 2017 warning on gabapentinoids with opioids.
• Full list with links: research thread 08, em-research/08-pain-side-effects-cognition.md.