The Surgery Is Not the Hard Part

Everyone will tell you to worry about the anesthesia. The thing that actually decides how this goes is the six weeks afterward.

Will the anesthesia make her dementia worse?

It may cause delirium, and delirium is associated with worse cognitive outcomes. Whether the anesthesia itself causes lasting decline is far less settled than the fear suggests.

Two different things get collapsed together here, and separating them helps.

Postoperative delirium is an acute confusional state in the days after surgery — disorientation, agitation or withdrawal, hallucinations, a sleep-wake cycle turned inside out. It is common in older surgical patients and substantially more common in people who already have dementia, who are among the highest-risk groups there is.

Long-term cognitive decline is the separate worry, and here the picture is murkier. Patients over 70 who developed delirium after elective non-cardiac surgery had about double the risk of cognitive decline at six months compared with those who did not.

That is an association, and it is worth stating honestly. Delirium may be doing damage, or it may be a marker for a brain that was already more vulnerable — the people who become delirious are, on average, the people with less reserve to begin with. Both are probably true to some degree. What follows practically is the same either way: preventing delirium is worth real effort.

Should we ask for a spinal instead of general?

This is the advice everyone gives, and the best randomized evidence does not support it as reliably as people assume.

The RAGA randomized trial compared regional with general anesthesia in older patients having hip fracture surgery and found no significant difference in postoperative delirium. Another trial in adults 70 and over having hip surgery found delirium in 9.5% of the general anesthesia group and 12.5% of the regional group — numerically the opposite of the expected direction.

Other work points the other way. A large meta-analysis did find higher delirium incidence with general anesthesia. And several systematic reviews have found no difference between the two in postoperative cognitive dysfunction at one week or three months.

Where the anesthesia conversation does matter is in the details — avoiding benzodiazepines for premedication, avoiding anticholinergic drugs, minimizing sedation depth, and treating pain well enough that pain itself does not drive the delirium. Those are worth asking about by name.

What is the question that actually matters?

Not "can she survive the operation." Almost certainly she can. The question is whether she can do the recovery, because that is where the benefit lives.

This is the part that goes unsaid, and it is the whole thing.

A knee or hip replacement is not a procedure that fixes a joint while you sleep. It is a procedure that creates the conditions for a good outcome, which are then achieved — or not — over weeks and months of physical therapy, adherence to weight-bearing restrictions, and daily exercises done at home when nobody is watching.

Every one of those requires the capacities dementia removes first: remembering instructions from an hour ago, understanding why something painful is necessary, sustaining effort toward a benefit weeks away, and inhibiting the impulse to get up and walk on a joint that is not ready.

Surgeons are, in general, honest about surgical risk. Far fewer will spontaneously walk a family through the rehabilitation demand, because it is not their part of the process. You have to ask.

What does the evidence show for joint replacement in dementia?

Higher complication rates and more discharges to nursing facilities — alongside real pain relief and largely satisfied families. It is not a straightforward no.

Reviews of joint replacement in patients with dementia consistently report higher complication rates, longer hospital stays, greater use of resources in the first 90 days, and more frequent discharge to skilled nursing or long-term care rather than home.

Specific risks run higher in the first year, including dislocation of the prosthesis and fractures around the implant — both of which are, in part, downstream of the same problem: an inability to reliably follow the movement restrictions that protect a new joint.

The reported in-hospital and 30-day mortality is often not significantly different. Longer-term mortality after hip surgery does appear worse in dementia, though the disease itself accounts for much of that.

And the other half of the finding deserves equal weight: the surgery relieves pain, and in these studies families report satisfaction with the result. Case series describe genuine difficulties — slow progress in therapy, getting out of bed unattended, pulling out IVs and catheters, agitation — and still conclude that the improvement in quality of life can outweigh them.

The evidence does not say don't. It says go in with your eyes open about what the hard part will be.

What makes recovery hard, specifically?

The hospital environment, the restrictions that cannot be remembered, and a therapy program built for someone who can carry instructions between sessions.

Worth picturing before you decide:

• Weight-bearing restrictions do not survive. "Toe-touch only for six weeks" is an instruction that requires continuous recall. She will stand up on it.
• Lines and drains get pulled. IVs, catheters, drains and dressings are unexplained objects attached to a confused person.
• Physical therapy assumes carryover. The therapist teaches exercises on Tuesday expecting them repeated Wednesday through Monday. Without that, progress stalls.
• The unfamiliar room is itself a risk factor. Hospitalization is hard on people with dementia even without an operation — the disruption, the sleep loss, the strange environment, the loss of routine.
• Pain reporting fails. Someone who cannot say "it hurts here" gets less analgesia, and untreated pain worsens delirium, which worsens everything else.
• Discharge is often not to home. Plan for that possibility in advance rather than discovering it on day three.

What reduces the risk if you go ahead?

Nearly all of it is arranged before the operation, not after.

The interventions with the best evidence are the unglamorous ones, and they are the same components that reduce delirium generally:

• Send the glasses and the hearing aids in, and insist they are used. A person who cannot see the room or hear the question will fail a confusion assessment and become genuinely more confused. This is the single most under-done intervention in inpatient care.
• Ask for a medication review before admission — specifically for anticholinergics and benzodiazepines, which raise delirium risk and are often still on the list.
• Ask that sleep be protected, and that overnight vital sign checks be clustered rather than scattered.
• Ask for early mobilization, and for hydration to be actively managed.
• Arrange for a familiar person to be present as much as the hospital permits, including in recovery. Familiarity is a real intervention, not a comfort.
• Bring a one-page sheet — what she responds to, what her baseline actually is, how she shows pain, what her name for you is. Staff cannot know what normal looks like for her, and without that they cannot recognize a change.
• Ask about a geriatrics or delirium-prevention consult. Many hospitals have one and will not offer it unless asked.
• Settle the discharge plan before surgery. Where will she go, who will supervise the restrictions, who will do the exercises with her, and is that person real and available.

What if we do not operate?

Then the pain has to be treated properly, because untreated pain is not the neutral option.

Declining surgery is not the same as doing nothing, and this is where families often drift.

Untreated pain in dementia drives agitation, resistance to care, poor sleep, reduced eating, immobility, deconditioning and falls. It is under-recognized precisely because the person cannot report it, and it frequently gets managed as a behavior instead of as pain. If the operation is off the table, the pain plan has to become an actual plan.

Reasonable things to ask about: scheduled rather than as-needed analgesia, a trial of adequate pain relief to see how much of the "behavior" resolves, physical therapy aimed at function rather than cure, injections, bracing, mobility aids, a home assessment for the specific tasks that hurt, and — where the trajectory warrants it — a palliative care referral, which is about symptom control and does not require anyone to be dying.

How do we actually decide?

By asking what this surgery is meant to achieve for this person, and how much of that survives if the rehabilitation does not happen.

Questions worth asking out loud:

• "What is the realistic best outcome for her specifically, and how long until she gets there?"
• "How much of the benefit depends on physical therapy she has to remember to do?"
• "What happens if she cannot follow the weight-bearing restrictions — what is the plan then?"
• "Where will she go when she is discharged, and for how long?"
• "What is the honest alternative if we do not do this — what does her pain look like in a year?"
• "Given her stage, what would you do if she were your mother?"

There is no general answer. A person in early-stage dementia with good family support and a knee that is stopping her walking is a different case from someone in stage six who cannot follow a two-step instruction. The decision turns on which of those is closer to your situation — and that is a judgment you are entitled to make with real information rather than reassurance.