The middle joint of each finger works every time you make a fist. When arthritis wears it out and non-surgical treatment no longer helps, two operations are usually considered. One is fusing the joint so that it no longer moves. The other is replacing it with an artificial joint. This guide explains how the two differ, who each one tends to suit, what can go wrong and what recovery involves. It also sets out what research has not yet settled.

The PIP joint and what damages it

Each finger except the thumb has three joints. The middle one is the , or PIP joint. It does much of the bending when you grip.
The most common cause of a worn PIP joint is osteoarthritis. Firm bony lumps called Bouchard nodes often form over the back of the joint. Symptoms, diagnosis and non-surgical care are covered in our guide to finger osteoarthritis. A PIP joint can also be damaged by an old injury, such as a fracture into the joint. Inflammatory arthritis, such as rheumatoid arthritis, is a third cause. The cause matters, because results of surgery can differ between these groups.

When is surgery considered?

Surgery is considered when non-surgical treatment has not helped enough. Pain, an unstable joint or stiffness still clearly gets in the way of daily life. A review published in 2018 found no trials comparing surgery with carrying on with non-surgical care1. The decision therefore rests on how much the finger troubles you, what you need your hand for and what you hope to gain.
Both operations are done for pain, instability or loss of function. They are not done only to change how a knobbly joint looks.

The two options in brief

  • Fusion aims to ease pain and give a stable joint by letting the bones heal together for good. That joint no longer bends, but the other finger joints still move.
  • Joint replacement swaps the worn joint for an artificial one. It keeps movement at the joint, usually about as much as before surgery rather than full movement.
In a study that compared the two operations in the index finger, pain relief and satisfaction were similar. Complications were more common after replacement. It was not a randomised trial. It looked at 65 replacements and 14 fusions, followed for a median of 72 and 8 months2. In other studies, movement after a silicone implant did not change much from before surgery34.

Fusion

In an , or fusion, the surgeon removes what is left of the cartilage from the two bone ends. The finger is set slightly bent, and the bones are held with wires, a screw or a wire loop until they join into one. Several methods are used, and each can lead to fusion. A review looked at 64 studies. The bone failed to heal in 3.9% of joints fixed with a compression screw and in 8.6% fixed with wiring passed through the bone5.
Fusion usually eases pain, but it may limit grip strength. Researchers tested this in healthy volunteers. They wore splints that held the PIP joint still at different angles, as a fusion would. Whole-hand grip was weaker at every angle tested. With the index finger held still, key pinch — the grip between the thumb and the side of the index finger — was also weaker6. For the index and little fingers, grip did not differ significantly between the angles tested6. In a study of the middle and ring fingers, the straighter the finger was held, the weaker the grip. Holding the middle finger still weakened grip more than holding the ring finger still7.
This is why the fusion angle is chosen for each finger and for how you use your hand. These splint studies were done in healthy people, so they show the direction of change, not the exact result after surgery. No study was found that used patient outcomes to identify one best angle.

Joint replacement

In , the worn joint surfaces are removed and an artificial joint is put in. There are two main types of implant: silicone and surface replacement.

Silicone implants

A silicone implant is a one-piece, flexible spacer that works like a hinge between the bones. In one study, people with osteoarthritis were seen about 10 years after surgery on average. Their pain was low. In a questionnaire they said they were satisfied and would choose the same operation again. They rated the look of the finger and the gain in function as neutral. Movement was no greater than before surgery3.
In the same study, X-rays showed that 21 of 38 implants had broken. Three needed revision surgery because a break caused symptoms, and one was removed because of infection. After an average of 10 years' follow-up, 90% of implants were still in place. The X-ray findings were not linked with how satisfied people were3. Only 22 of 51 people came back for review, which limits how certain these results are.
A review of 34 studies found that silicone implants least often needed a further operation in which the implant was not changed. The figures were 1% for silicone, 7% for pyrocarbon and 10% for metal-and-plastic implants. Revision surgery, an operation that changes the implant, was needed at similar rates for all types, 2% to 4%. A finger tilted to one side or an unstable joint was reported more often after silicone implants. The authors saw silicone as a good option mainly for stable joints8. Another review, limited to osteoarthritis, found silicone implants had the lowest pain scores and fewest complications. How long the implants lasted did not differ between types9.

Surface replacement

A surface replacement replaces the worn ends of both bones that form the joint. It has two parts fixed into the bones, which glide on each other much like a natural joint. These implants are made of different materials, such as metal with plastic, or pyrocarbon8.
One of the surface replacements that has been studied is CapFlex-PIP. In a small study, its parts had become bonded to the bone at one year10. In another study, 92 were implanted; at 5 years, 68 implants in 65 people were available for review. In that group, pain during daily tasks fell on average from 6.4 to 1.8 on a 0 to 10 scale. Movement at the joint rose on average from 45° to 54°. Before surgery, 65% of the fingers were tilted to one side by more than 5°; at 5 years, 25% were. Eight people needed a further soft-tissue operation, and 4 of the 92 implants needed revision surgery for stiffness or loosening11.
The same centre's registry compared 279 CapFlex-PIP surface replacements with 424 silicone implants. Complications were more common after surface replacement (20% versus 11%). So were further operations without changing the implant (5.4% versus 0.5%). Revision surgery, an operation that changes or removes the implant, did not differ significantly (4.4% versus 3.3%). Surface replacement less often left the finger tilted to one side (19% versus 58%)12. The authors suggest it mainly for fingers that are tilted but can be corrected, accepting a higher chance of a second operation12. In a registry of 476 of these implants, more surgical experience was linked with fewer implant problems and less revision surgery. The little finger had implant problems more often13.
A large part of the CapFlex-PIP data cited here come from one centre in Switzerland, so how far the results apply elsewhere is limited. No randomised trial comparing it with a silicone implant was found. It cannot be said to be better than other implants.
Pyrocarbon surface replacements have been compared with silicone implants in a review. Pain relief and movement were similar, but further operations for complications were more common after pyrocarbon14.

How do the two types compare?

Comparison is difficult. Of 50 papers on PIP joint replacement published from 2010 to 2021, 82% were case series and only one was a randomised trial. There is no agreement on which implant is best15. A randomised trial in three centres included 62 joints in 43 people. It compared a silicone implant with two surface replacements, one titanium-and-plastic and one pyrocarbon, not CapFlex-PIP. Pain fell in all groups, but the surface replacements had to be removed more often16. Studies differ in implant type, cause of arthritis, surgical approach and length of follow-up. Figures from different studies should not be set side by side as if they were comparable.
With that caution, the evidence suggests:
The findings are not consistent. One review found revision surgery after 18% of surface replacements put in from the back of the finger. For silicone implants put in from the palm side, the figure was 6%. This was not a direct comparison. The implant, the approach and the length of follow-up all differed17.
The joint can be reached from the back of the finger or from the palm side. In a review, movement and patient-rated results were similar with both approaches. A — the middle joint bending backwards while the end joint bends down — developed after some back approaches (9 of 82 implants). None was seen after the palm-side approach (0 of 101)18. Your surgeon will discuss which approach suits your finger.

Which finger?

When you turn a key, your thumb presses on the side of the index finger. An artificial joint there has to resist strong sideways force. A review of five small studies looked at problems linked to an unstable joint after replacement. They were found in 33% of index fingers and 29% of middle fingers. In ring and little fingers the figure was 6%. The studies used different implants and defined these problems differently19.
In the same observational study of the index finger, both types of pinch tested got stronger after fusion, but only one after replacement. Pain relief and satisfaction were similar, and the risk of complications was 4.3 times higher after replacement2. It is common teaching that replacement suits the index finger less well than fusion19. This rule rests on few studies, not on randomised trials. In the same review, the middle finger was not shown to be at lower risk of these problems than the index finger19.
Fusion has a cost that also depends on the finger. Holding the middle finger still weakened grip more than holding the ring finger still7. When several fingers are affected, the plan is made for the whole hand rather than one joint at a time.

Stable or deformed joint?

The choice also depends on the state of the joint. How much bone there is and how strong it is matter. So do working tendons and ligaments. With major bone loss or instability that cannot be corrected, a joint replacement may not be suitable and fusion is considered2021.
  • Stable, straight joint. Either operation may suit. A review of complications saw silicone implants as a good option for stable joints8.
  • Finger tilted to one side but still correctable. In the registry cited, CapFlex-PIP kept the finger straighter than silicone implants, although a second operation was more often needed12. Authors of a review of complications consider that surface replacements might suit these fingers better8.
  • Fixed finger deformity. A is the middle joint bending down while the end joint bends back. In a long-term study of silicone implants, correcting this or a swan-neck deformity was difficult and usually gave poor results4. A clinical review advises fusion rather than replacement for these fixed deformities21.

Osteoarthritis, injury or inflammatory arthritis?

Most studies are about osteoarthritis. Many also include joints damaged by injury21917. In one study, silicone implants were used for several causes. People with rheumatoid arthritis had less pain relief and less movement than people with osteoarthritis or an old injury4. Studies of some surface replacements included people with rheumatoid or psoriatic arthritis, but they do not allow firm conclusions for each of these groups22. Results from osteoarthritis studies may not apply in the same way to these conditions.

What can go wrong?

After PIP joint replacement, problems can include:
After fusion, an important risk is that the bones do not heal together5. Permanent loss of movement at the fused joint is intended, but it may affect grip strength6.

What if a joint replacement fails?

An implant can be replaced or changed in revision surgery, but the results are less predictable. In a series of 75 revision operations, 19 fingers needed a second revision. The estimated chance of avoiding one by 5 years was 70%. Complications were common, and an unstable joint was linked with worse results. Most people had good pain relief but less movement23. After a failed silicone implant, a repeat operation helped most when the finger was very stiff. A sideways tilt could not be fully corrected24.
Another option is to remove the implant and fuse the joint. This can be difficult. In a small series, the bones healed in 8 of 13 fingers, and people rated their results as fair to good25.

Denervation

Cutting the small nerves that carry pain from the joint has also been proposed. The joint stays in place and keeps its movement. A review covering several hand joints found that side effects were common, mainly changes in feeling and the need for more surgery. The evidence did not allow any conclusion about whether it works26. A study from two centres covered 116 fingers with PIP joint arthritis. The estimated chance of avoiding a salvage operation by 10 years was 82%27. That study looked back at past cases and had no comparison group.

Recovery and hand therapy

Hand therapy plays a large part after either operation. A teaches exercises, fits splints and helps keep the other finger joints from stiffening.
After joint replacement, most published programmes include exercises to move the joint. The finger is protected with splints, either rigid or dynamic, or by taping it to the next finger. Programmes vary between centres, and there is no evidence for one best plan28. After fusion, the finger is protected until X-rays show the bones have healed.
Your surgical team will set your exact plan, including how long to wear a splint and when to return to work or driving. This depends on the type of operation, the implant and how it was done. General advice on looking after your hand is in our guide to recovering after hand surgery.

Questions to discuss with your surgeon

  • What matters most to you: pain relief, a stable and strong finger, or keeping movement?
  • Which finger is affected, and what do you most need it for?
  • Is the joint stable and straight, or tilted or deformed?
  • What caused the damage: osteoarthritis, injury or inflammatory arthritis?
  • What is the risk of another operation, and what can be done if an implant fails?
  • How long will hand therapy take, and can you fit it around work and other commitments?
  • How much experience does the surgical team have with the type of implant offered?

When to seek help after surgery

Contact your surgical team or doctor promptly if:
  • redness, warmth, swelling or pain around the wound is getting worse, fluid leaks from the wound, or you have a fever
  • the finger turns pale, blue or numb, or the splint or dressing feels too tight
  • the finger suddenly changes position, clicks out of place or feels unstable
  • severe pain comes back after a period of improvement
If the finger is pale, blue or cold, or loses feeling, act quickly. If you cannot reach your surgical team, go to a hospital emergency department straight away.

Summary

When a PIP joint is worn out and non-surgical care is not enough, it can be fused or replaced. Fusion usually eases pain and gives a stable joint, but that joint permanently stops bending. Replacement keeps movement, usually about as much as before surgery. In the index-finger study cited, complications were more common after replacement than after fusion. Silicone implants are a good option for stable joints in particular. In the registry cited, CapFlex-PIP kept the finger straighter than silicone implants but more often needed a second operation. A large part of its data come from one centre. No single implant is best for everyone. In the index finger, fusion is traditionally preferred, though this rests on few studies. Few studies compare these options directly. The choice is made together with your surgeon. It takes into account the finger, the state of the joint, the cause of the arthritis and what you need your hand to do.