What Happens If a Cortisone Shot in the Shoulder Doesn’t Work?

When a cortisone shot in your shoulder fails to bring relief, the usual next step is a diagnostic reassessment followed by a shift to alternative treatments. A single failed injection does not mean you are out of options, but it does change the conversation with your doctor. The reasons cortisone can fall short range from an inaccurate diagnosis to the shot simply landing in the wrong spot, and each reason points toward a different path forward.

How Cortisone Is Supposed to Work in the Shoulder

Understanding why cortisone fails starts with knowing what it actually does when it succeeds. Corticosteroids work by blocking the enzymes that produce inflammatory chemicals in your tissues, which reduces swelling and pain at the site of injection. A local anesthetic mixed in with the steroid provides immediate but temporary numbness, letting you move the shoulder sooner and avoid the stiffness that can make inflammation worse over time.¹ That relief typically peaks in the first couple of weeks and can last anywhere from a few weeks to a few months. When it works, the window of reduced pain is supposed to let you do physical therapy more effectively, addressing the mechanical problem underneath the inflammation.

The shot is not a cure for whatever is causing your shoulder pain. It is a tool to temporarily quiet the inflammation so that rehabilitation or natural healing can take place. When that temporary window closes and the pain returns unchanged, or when the shot never provides relief in the first place, it tells your doctor something important about what is going on inside the joint.

Why the Injection Might Not Have Worked

There are several distinct reasons a cortisone shot can fail, and they do not all lead to the same conclusion.

The most straightforward explanation is that the needle did not reach the right structure. The shoulder is a complex joint with multiple potential pain sources packed into a small area. A shot aimed at the subacromial bursa, for instance, can miss and end up in surrounding tissue. A systematic review comparing ultrasound-guided injections with “blind” (landmark-guided) injections found that the guided approach produced meaningfully greater pain improvement and functional gains at six weeks.² If your first injection was done without imaging guidance, a repeat attempt using ultrasound may reach the target more reliably. That said, one randomized trial found both approaches had comparable outcomes, suggesting that accuracy alone does not explain every failure.³

A second reason is that the diagnosis itself may be incomplete or wrong. You might have been treated for subacromial bursitis or rotator cuff tendinopathy when the real source of pain is adhesive capsulitis (frozen shoulder), a labral tear, osteoarthritis of the glenohumeral joint, or a problem originating in the neck rather than the shoulder. These conditions overlap considerably in how they feel, and a cortisone shot targeted at bursitis will not fix a torn labrum or cervical nerve compression. If the injection provided zero relief, not even a brief improvement from the local anesthetic, that is a strong hint that the pain generator was not where the needle went.

A third possibility is that the underlying structural damage is simply too advanced for cortisone to make a dent. A large rotator cuff tear, for instance, is a mechanical failure. Reducing inflammation around it might ease some pain temporarily, but it will not restore the torn tissue’s ability to stabilize the joint. Similarly, significant cartilage loss from shoulder arthritis produces pain that responds poorly to cortisone alone.

What Your Doctor Will Likely Reassess

A failed cortisone shot usually prompts your doctor to reconsider the diagnosis. If you have not had advanced imaging, this is typically when an MRI or diagnostic ultrasound gets ordered. These can reveal structural problems that a physical exam alone might miss, such as partial-thickness rotator cuff tears, labral damage, or calcium deposits within a tendon.

Your doctor may also reconsider the possibility that the pain has a neurological component. The suprascapular nerve, which carries roughly 70 percent of the shoulder’s sensory signals, can become compressed or irritated, producing deep shoulder pain that mimics rotator cuff problems but does not respond to local cortisone.⁴ When nerve involvement is suspected, targeted nerve blocks can serve a dual purpose: they confirm the nerve is the pain source and provide relief at the same time.

Another possibility that gets explored after a failed injection is whether your pain has become centrally sensitized. In chronic shoulder pain, the nervous system can amplify pain signals so that the brain perceives more pain than the tissue damage alone would produce. Research has found that patients whose shoulder pain has features of central sensitization tend to have significantly worse function and higher levels of anxiety, depression, and catastrophizing about their pain.⁵ A hallmark sign is heightened pain sensitivity that extends beyond the shoulder itself, sometimes affecting the entire upper limb.⁶ Cortisone injected into the shoulder will not quiet a pain signal that is being amplified in the spinal cord and brain, which is why recognizing this pattern matters: the treatment shifts toward the nervous system rather than the shoulder joint.

Physical Therapy After a Failed Shot

Even when cortisone does not deliver the relief you hoped for, structured physical therapy remains a cornerstone of shoulder treatment. A large pragmatic trial of patients with moderate to severe shoulder pain from impingement found that about a third of adults did not respond adequately to exercise therapy alone within three months and went on to receive an injection.⁷ But the reverse scenario, where the injection fails and exercise becomes the primary treatment, is equally common. The logic is straightforward: many shoulder conditions involve weakness and poor movement patterns that no injection can fix, and the strengthening and mobility work you do in therapy addresses those mechanical deficits directly.

When pain is a barrier to doing the exercises, your doctor may pair physical therapy with other pain-management strategies. For adhesive capsulitis (frozen shoulder), one approach that has shown promise is combining a nerve block of the suprascapular nerve with an intra-articular corticosteroid injection. The nerve block helps control the intense pain of the most debilitating stage, allowing patients to participate more fully in therapy.⁸ The principle here is broader than one study: if a single injection method did not buy you enough pain relief to rehabilitate the shoulder, your doctor may try a different injection target or technique alongside therapy rather than abandoning injections entirely.

Platelet-Rich Plasma and Other Injectable Alternatives

Platelet-rich plasma (PRP) is probably the most widely discussed alternative injection after cortisone fails. PRP is made from your own blood, concentrated to increase the platelet count, and injected back into the injured area. The idea is that the growth factors in platelets promote tissue healing rather than just suppressing inflammation. How does it actually compare?

A systematic review and meta-analysis of PRP versus cortisone for rotator cuff injuries found that PRP was somewhat better at reducing pain in the short term, but the two were similar at later time points. Where PRP showed a more noticeable edge was in improving function over time: the functional gap between PRP and cortisone widened as the months went on.⁹ A separate study following patients for up to 18 months found no significant differences between PRP and cortisone in pain scores, shoulder function, or range of motion at any time point.¹⁰

The honest reading of the current evidence is that PRP is not a dramatic upgrade over cortisone for most rotator cuff problems. If cortisone failed because the underlying structural damage is too severe, PRP is unlikely to succeed either. Where PRP may offer a genuine advantage is in situations where you want to avoid the tissue-weakening effects of repeated cortisone (more on that below) or where a longer-term functional benefit matters more than fast pain relief.

For shoulder osteoarthritis specifically, hyaluronic acid injections are sometimes tried when patients have already failed conventional treatments including cortisone. A prospective trial of hyaluronic acid in patients who had not responded to anti-inflammatory drugs, at least one steroid injection, and a month or more of physical therapy found that it could still provide some improvement in that population.¹¹ Hyaluronic acid acts as a lubricant and cushion inside the joint rather than an anti-inflammatory, so it is working through a different mechanism entirely. The evidence for it in shoulders is thinner than for knees, but it is an option your doctor may raise.

When Calcium Deposits Are the Problem

Calcific tendinopathy, where calcium deposits form inside the rotator cuff tendons, is a condition that frequently does not respond well to cortisone injections alone. The deposits cause mechanical irritation and intense inflammatory flares, and while cortisone can calm the inflammation temporarily, it does nothing about the calcium sitting inside the tendon.

When cortisone and other conservative treatments fail for calcific tendinopathy, a procedure called ultrasound-guided barbotage is often the next step before surgery. In barbotage, a needle is inserted under ultrasound guidance to break up and aspirate the calcium deposit.¹² Extracorporeal shock wave therapy is another option used when initial treatments have failed but before turning to surgery.¹³ Ultrasound-guided interventions like barbotage and percutaneous ultrasonic tenotomy have emerged as effective minimally invasive alternatives to open surgery for cases that do not respond to conservative care.¹⁴

One study comparing post-needling injection with PRP versus cortisone for calcific tendinopathy found that PRP led to greater pain reduction at 12 months in harder, more formed calcium deposits, though improvements in function and complete resorption rates were similar between the two.¹⁵ The takeaway here is that calcific tendinopathy has its own treatment ladder, and a failed cortisone shot moves you up that ladder, not to the end of it.

Why You Should Not Just Keep Getting More Shots

When one cortisone shot does not work, a natural impulse is to try another. Sometimes that is reasonable, particularly if the first injection may have missed its target and a second attempt with ultrasound guidance could do better. But stacking cortisone injections has real downsides that increase with each round.

Basic science research has documented what corticosteroids do to tendon tissue. Exposure changes the normal structure and alignment of tendon cells, decreases the ratio of strong structural collagen to weaker collagen, and causes changes in cell shape and orientation. After high-dose exposure, the tissue’s appearance did not return to normal even three weeks later.¹⁶ Repeated injections compound this damage, potentially weakening the very tendons you are trying to protect. This is the main reason most orthopedic specialists limit cortisone shots to three or four in the same joint per year, and many prefer fewer.

There is also a practical timing issue if surgery eventually becomes necessary. A systematic review found that patients who received a shoulder cortisone injection within three months before shoulder arthroplasty (joint replacement) had a markedly higher infection rate.¹⁷ For arthroscopic procedures, the risk window is especially steep in the first two weeks. Patients who received cortisone within two weeks of arthroscopic shoulder surgery had roughly three times the odds of postoperative infection at 90 days, and those injected two to four weeks before surgery still had more than double the odds. The elevated risk effectively disappeared when the injection was given more than four weeks before surgery.¹⁸ If there is any chance you will need shoulder surgery, the timing of your last cortisone injection matters a great deal, and getting an extra “just in case” shot close to a potential surgical date can create unnecessary risk.

Cortisone and Blood Sugar

If you have type 2 diabetes, a cortisone injection does not stay local in the way you might assume. A study tracking continuous glucose monitor data in patients with type 2 diabetes who received a shoulder cortisone injection found that average blood glucose jumped from about 136 mg/dL before the injection to 159 mg/dL the next day. The time spent at dangerously high levels (above 250 mg/dL) more than doubled on day one, from about 4 percent to nearly 10 percent.¹⁹ The good news is that glucose levels returned to baseline by about day three to four after injection. But if you are considering a second or third cortisone shot that is not providing lasting shoulder relief, the repeated glucose spikes become an additional reason to explore non-steroidal alternatives instead. Your endocrinologist and orthopedist should be communicating about this, but it often falls to you to make sure they are.

Nerve-Targeted Procedures for Persistent Pain

When shoulder pain persists despite cortisone, physical therapy, and other conservative measures, and surgery is either not indicated or not desired, nerve-targeted procedures represent a less commonly discussed option. Pulsed radiofrequency treatment of the suprascapular nerve and thermal radiofrequency ablation of the shoulder’s articular nerve branches have both been studied for intractable shoulder pain. In a randomized controlled trial, both approaches produced significant decreases in pain scores at one, three, and six months compared to baseline.⁴ These procedures work by interrupting pain signal transmission rather than reducing inflammation or healing tissue. They do not fix the underlying structural problem, but for patients who are not surgical candidates, or who are trying to delay surgery, they can provide months of relief that cortisone could not.

When Surgery Enters the Picture

Surgery is generally considered only after a thorough trial of non-operative treatments has failed. For rotator cuff tears, that trial typically includes physical therapy, at least one cortisone injection, and often a PRP or other injection attempt. For calcific tendinopathy, surgery to remove the calcium deposit and repair any associated rotator cuff damage is used once barbotage, shock wave therapy, and conservative measures have all been exhausted.¹³ For severe shoulder osteoarthritis, the endpoint may be joint replacement. For large or complete rotator cuff tears, surgical repair or tendon transfer procedures become the discussion.

The timing question matters here. If your failed cortisone shot leaves you in a position where surgery looks likely, give your body at least four to six weeks between the last injection and the procedure, based on the infection risk data described earlier. Some surgeons prefer a longer buffer. If you had an injection recently and are now being referred to a surgeon, bring the date of your last shot to that consultation so it can factor into scheduling.

One thing worth knowing is that a failed cortisone shot does not mean surgery will also fail. The two work through entirely different mechanisms. A cortisone injection treats symptoms; surgery addresses the structural cause. Plenty of patients who got no benefit from cortisone go on to have excellent surgical outcomes, particularly when the diagnosis has been refined and the underlying problem clearly identified before the operation takes place.