Arthrocentesis and Injection of Joints and Soft Tissue

54 Arthrocentesis and Injection of Joints and Soft Tissue

Arthrocentesis and injection of joints are safe and simple procedures that can be performed routinely at an outpatient visit.1 With analysis of synovial fluid, few procedures in medical practice have the potential to be as diagnostically definitive as arthrocentesis and few modalities can be as effective in achieving symptomatic relief of painful or swollen articular structures as the injection of corticosteroids. For these reasons, one out of five visits to a rheumatology practice includes aspiration or injection of a joint or periarticular structure. However, a majority of internists finishing their residency training believe they need more training in these important, safe, and effective procedures, and most injections performed in primary care settings are done by a small percentage of practitioners with experience and comfort with the procedure. Recent efforts to formalize educational processes in arthrocentesis and joint injection have the potential to increase the number of primary care physicians who use these procedures more regularly.

Paracelsus is credited with the first descriptions, in the early sixteenth century, of the viscous fluid present within synovial cavities, but aspiration of synovial fluid for analysis and aid in diagnosis did not become a topic of increasing interest until the first half of the twentieth century. Numerous studies of synovial fluid components and techniques used to obtain this fluid appear in a 1935 textbook by Pemberton.2 An early description of arthrocentesis technique can be found in the classic book by Ropes and Bauer3 on synovial fluid analysis, published in 1953, which used data collected over a 20-year period. During this era, swollen, distended joints were often aspirated for relief of discomfort. In most instances, the prompt reaccumulation of fluid and concerns about infection from repeated aspirations limited the usefulness of aspiration for the relief of arthritis symptoms. A wide variety of substances were injected into joints throughout the early twentieth century including formalin and glycerin, ethiodized oil (Lipiodol), lactic acid, petroleum jelly, and liquefied oil prepared from the patient’s own subcutaneous fat. Most of these therapies were apparently abandoned, and the most therapeutic injections discussed in arthritis textbooks from the 1940s related to temporary relief via injections of procaine for osteoarthritic knees and bursitis of the shoulder.

In 1951, after observations on the efficacy of topical cortisone for ocular inflammation, Hollander first reported a minimal, transient improvement in 25 knees of patients with rheumatoid arthritis when injected with cortisone. In subsequent years, Hollander injected hydrocortisone acetate with a much better response, providing further evidence that this was the active anti-inflammatory metabolite of cortisone.4 Further reports of benefit from injectable corticosteroids appeared in the 1950s. During this period, studies showed that more stable and less soluble compounds in the form of esterified crystalline hydroxycortisone and its analogues were even more effective and had longer duration of anti-inflammatory effects. By the early 1960s, Hollander had reported a series of more than 100,000 injections of joints, bursae, and tendon sheaths in 4000 patients with a variety of conditions. In rheumatology practice since then, aspiration and therapeutic injection of joints and periarticular tissues have become common and essential procedures.

Indications and Clinical Evidence


Aspiration of synovial fluid may be indicated in any joint with detectable effusion or may be attempted in joints without detectable effusions when diagnosis is in doubt (Table 54-1).5 In patients in whom a diagnosis is uncertain, synovial fluid analysis usually provides important information regarding the inflammatory or noninflammatory nature of the process within the affected joint and may be definitive in patients with crystal-induced or infectious arthritis. In patients with recently diagnosed bacterial arthritis, repeated aspiration of accumulated fluid is often an important adjunct to antibiotic therapy. Joints with detectable effusions may be aspirated for relief of discomfort, with or without a subsequent injection of corticosteroid. In some patients, aspiration alone, without steroid injection, may be particularly effective for noninflammatory effusions or self-limited conditions.

Table 54-1 Indications for Arthrocentesis

Undiagnosed Arthritis with Effusion

Undiagnosed Arthritis without Effusion

Patient with Known Diagnosis

WBC, white blood cells.

* Most studies show improved effect if fluid aspirated before injection.

Therapeutic Injection

Inflammatory Arthritis

Therapeutic injection of corticosteroids is generally believed to be most effective in joints affected by inflammatory arthritis (Table 54-2). Most of the experience in this area has been with common conditions such as rheumatoid arthritis, juvenile rheumatoid arthritis, crystal-induced arthritis, psoriatic arthritis, and reactive arthritis. Anecdotal experience has been reported in less common conditions such as systemic lupus erythematosus and sarcoidosis. In self-limited conditions such as gout, injections generally lead to more prompt resolution of exacerbations.

Table 54-2 Indications for Therapeutic Injection

Inflammatory Arthritis

Noninflammatory Arthritis

Nonarticular Conditions (Tendinitis, Bursitis, Myofascial Pain)

In rheumatoid arthritis, injections are used frequently to suppress inflammation in individual joints. Such injections are generally considered to be adjunctive to disease-modifying drug therapy and are not believed to affect overall outcomes. The efficacy of individual joint injections in rheumatoid arthritis is supported by many large, uncontrolled case series dating back to the 1950s. Most of Hollander’s early reports suggested long-lasting relief in most joints injected, with improvement lasting several months in most patients.6 This long-lasting relief has been confirmed in several series of patients in subsequent years. In 1972 McCarty7 reported that 88% of patients attained remission for an average of 22 months in small joints of the hands and wrists, much better than comparable joints not injected in the opposite hands of the same patients. In a subsequent report on 956 injections in 140 patients followed for an average of 7 years, 75% of injected joints remained in remission.8 In this series, patients received about two injections during the first year of treatment and averaged 0.6 injection per patient-year for the next 15 years.

More recently, multiple intra-articular injections in inflamed joints have been shown to be a useful part of an overall regimen of disease-modifying therapy in rheumatoid arthritis, resulting in improvement superior to similar doses of systemic steroids, and helpful in obtaining clinical remission and reducing radiographic progression of disease.911 In particular, intra-articular injections of steroids added to systemic methotrexate therapy have been shown to improve periarticular osteopenia in swollen hand joints over a 3-month period in patients with rheumatoid arthritis.12 Removal of fluid before aspiration increases the efficacy of steroid injections in most patients with inflammatory arthritis. One study of 191 knee injections in patients with rheumatoid arthritis showed that aspiration of fluid reduced the rate of relapse from 47% to 23% within a 6-month period after injection compared with joints not aspirated.13

Corticosteroid injection is considered to be a safe and effective option for prompt relief of acute crystal-induced arthritis in gout and pseudogout. Steroid injections are so widely accepted as an effective treatment that few reports have attempted to address the degree or duration of efficacy for injections in these conditions. Many of the patients described in early reports of steroid injections were being treated for acute crystal-induced arthritis, with prompt relief being almost uniform. In a more recent report, small doses of intra-articular steroids were successful in relieving pain and swelling completely in all patients within 48 hours, and no relapses were noted in 20 patients over a 3-month period.14

In patients with recent onset of inflammatory oligoarthritis, without definitive diagnosis, corticosteroids can be used to relieve symptoms of swelling in individual joints, and the response to these injections can be used as a prognostic marker. In a series of 51 patients with recent-onset inflammatory arthritis involving five or fewer joints, a strategy of injecting all joints with clinical synovitis resulted in improvement in all patients and a complete resolution of synovitis after 2 weeks in 57% of patients. The response at 2 weeks was the best predictor of continued improvement persisting for 26 weeks and 52 weeks.15 A trial of steroid injection may be used for symptomatic relief in such patients, and a complete response at 2 weeks can be used as a prognostic indicator for a better outcome. Local injections are effective in 41% and 25% of joints at 3 and 12 months, respectively, in patients with localized flares of psoriatic arthritis, with the response more prolonged in patients taking effective long-acting systemic therapy.16

Patients with refractory sacroiliac joint pain related to ankylosing spondylitis or other spondyloarthropathies may benefit from injection of the sacroiliac joints.17 Because of the anatomy of this joint, such injections often require radiographic confirmation of needle placement in the joint space, and use of fluoroscopy-guided, computed tomography (CT)–guided, and magnetic resonance imaging (MRI)–guided injections has been reported. In uncontrolled studies, a good response has been reported in about 80% of injections, with an average time of improvement of 6 to 9 months. At least one controlled study in a small group of patients showed slight benefit of steroid compared with placebo injection. The degree of improvement seen after sacroiliac injections has not been consistent among studies to date, however, probably because of a lack of uniformity of patient selection and outcomes assessed.

Joint injection has been used with increasing frequency in recent years in patients with juvenile rheumatoid arthritis,18 particularly in the pauciarticular variant of the disease, in which only a few joints are involved, and potentially toxic systemic therapy can be avoided. Complete remission lasting more than 6 months has been reported in approximately 65% to 80% of joints injected in this condition, most commonly in the knees. Benefit has also been shown in smaller numbers of ankles, wrists, shoulders, elbows, and temporomandibular joints, with most children being able to stop oral medications, and correction of joint contracture being noted in most as well.19 A median duration of improvement of approximately 74 weeks has been documented in another large study. Joint lavage before injection may be useful in prolonging response in patients with a poor response to previous injection.20 One study showed a significant decrease in leg-length discrepancy in children treated with repeated injections (average of 3.25 injections per child over 42 months) compared with children in another center who were not injected.21 A recent decision analysis model suggests that intra-articular injection is superior to a strategy of initial nonsteroidal anti-inflammatory (NSAID) use in patients with monoarthritis of the knee.22

A more recent study specifically addressed the efficacy and safety of steroid injections in the hip in juvenile rheumatoid arthritis.23 In this prospective study of 67 hip injections, 58% of hips remained in remission for 2 years after a single injection; another 18% required a second injection to maintain remission. Only two cases of avascular necrosis were seen in this group, and both of these were in patients receiving systemic steroids, suggesting no role for local steroid injections in the development of avascular necrosis in this population.

Finally, local steroid injection may be a useful adjunct in managing patients with hemophilic arthropathy.24 In an open trial of 19 injections, 79% of joints improved within 24 hours; this improvement persisted for 8 weeks in 58% of joints. A decrease in need for clotting factor was shown in this small group.

Noninflammatory Arthritis

Corticosteroid injection is used frequently in common noninflammatory articular conditions such as osteoarthritis, internal derangements, and post-traumatic arthritis. Clinical studies that support efficacy are less convincing and suggest a less predictable and smaller degree of response in these conditions than is seen in inflammatory arthritis. Most studies of steroid injections in osteoarthritis have studied patients undergoing knee injections.25 Early uncontrolled studies suggested improvement in approximately 60% to 80% of patients.6 In controlled studies, most benefit, compared with placebo, seems to last 1 to 6 weeks, with return to the same pain levels seen in placebo groups by around 12 weeks after injection.2629

Factors associated with a better response to steroid injections have included less severe radiographic changes, the presence of effusion at the time of injection, and successful aspiration of fluid at the time of injection. The theoretic concern about the potential for negative effects of injected steroids on cartilage (discussed in the following section) is often cited as a reason to limit injections in osteoarthritis and other forms of noninflammatory arthritis. A more recent trial in 68 patients comparing corticosteroid injections every 3 months with saline injections showed no worsening of radiographic changes, however, after 2 years of repeated steroid injection, along with significant improvement in symptoms during the period of study.30

Injections in patients with osteoarthritis of the hip are less likely to be helpful and are more technically difficult. Some relief can be obtained, however, in patients with less severe disease or in a rare patient with more severe involvement. A prospective, open study of intra-articular steroid in 45 patients with hip arthritis, 27 of whom had osteoarthritis, found a significant reduction in pain at 2 weeks and 12 weeks, although the effect was lost by 26 weeks.31 In a report of 510 patients treated with a single injection done under fluoroscopic guidance, pain relief that persisted 8 weeks was seen in 90% of patients with mild disease, 58% of patients with moderate disease, and 9% of patients with severe hip osteoarthritis; improved range of motion was shown in most of the patients responding.32 Controlled trials have shown that steroid injection provides modest benefit compared with placebo for 2 to 12 weeks, and the benefit was no longer apparent at 3 months.33,34 Steroid injection followed by non–weight bearing was not helpful in reducing the need for hip replacement in a retrospective study of patients with rapidly progressive osteoarthritis of the hip.35 In patients with osteoarthritis of the thumb, local steroid injection may be helpful for 1 year in a few patients (≈20%), but most patients are improved for 1 to 3 months at the most.36

Injectable hyaluronic acid derivatives have been studied extensively and are frequently used for injection into osteoarthritic knees and occasionally in other joints. A series of one to five weekly injections has been shown to provide more pain relief than placebo in most studies. The degree and duration of improvement in these studies have varied, however, and the optimal role for hyaluronic acid injections in the management of arthritis has yet to be determined (see later discussion and Chapter 100).37,38

Nonarticular Conditions

Patients with various forms of tendinitis, bursitis, myofascial pain, and nerve entrapment syndromes are frequently treated with local injections of corticosteroids.39 In many of these conditions, uncontrolled clinical experience suggests a high response rate, and in many others, controlled trials show variable levels of benefit, often depending on whether short-term or long-term outcomes are considered. The injection of trigger points for pain relief has been used by many practitioners over the past several decades, but few controlled studies to support efficacy have been published.

Steroid injections are frequently used in the management of rotator cuff tendinitis, frozen shoulder, and other causes of shoulder pain. Most controlled studies have shown significant short-term improvement from steroid injection compared with placebo injection, usually lasting 4 to 6 months.40 In most such trials, short-term treatment success of 75% to 80% is usually reported in steroid-treated groups compared with 40% to 50% in placebo groups. Most, but not all, controlled studies have shown that steroid injections are superior to physical therapy without injection, and that combining steroid injection with physical therapy has an additive benefit.4143 In a subset of patients with painful shoulder related to calcific tendinitis, local injection of ethylenediamine tetra-acetic acid (EDTA) may result in pain relief and radiographic resolution of calcification.44 Local injections may also be useful for shoulder pain related to the acromioclavicular joint.

Lateral epicondylitis is also commonly treated by local injections.40 Pain may worsen for 1 or 2 days after injection but usually improves after 4 to 5 days.45 Longer controlled studies typically document improvement of 90% compared with 50% in placebo treatment in the first 1 or 2 months after injection, but outcomes at 6 to 12 months are usually not affected and one study has shown that recurrences of pain after 6 weeks were more common in patients who had steroid injections.46,47 Local injection of botulinum toxin has been shown to be beneficial compared with placebo for epicondylitis in small trials.48 In femoral trochanteric pain syndromes (i.e., bursitis), the response rate to locally injected steroids has been reported to be 60% to 100%, but no placebo-controlled studies have been done. A prospective study reported significant improvement after a single injection in 77% of patients at 1 week; this number decreased to 69% at 6 weeks and 61% at 26 weeks.49 Patients receiving larger amounts of locally injected steroid (24 mg of betamethasone) were more likely to have sustained improvement. Around the knee, anecdotal and retrospective studies have shown that most patients with anserine bursitis respond to local steroid injection.

Local steroid injection may be a useful nonsurgical therapy for carpal tunnel syndrome. Most studies report 90% short-term relief of symptoms from a single injection; longer-term relief is 20% to 90%, and surgery is eventually required in about half of patients treated with injection. Controlled trials comparing local steroid injection with surgical decompression have shown variable results, suggesting that injection may provide better relief within the first 3 to 6 months, but more patients benefit from surgery when followed for 6 to 12 months.50,51 A good response to local injection is sometimes useful as a diagnostic test and is a predictor of good surgical response. As noted later, care should be taken to avoid injection into the body of the median nerve.52

Most patients with de Quervain’s tenosynovitis involving the tendons at the base of the thumb respond to local steroid injection. In three prospective studies, 60% to 76% of patients with this condition had their symptoms adequately controlled with a single injection, and another 10% to 33% required a second injection.53,54 About 30% had exacerbations an average of 1 year later, but overall, only 10% to 17% of patients were not controlled and required surgical release. Another small controlled study showed that injection was much better than splinting.55 In patients with flexor tenosynovitis, steroids are effective in 88% compared with 36% of patients receiving saline injections, with improvement lasting for up to a year in most.56 Similar success rates have been reported in prospective studies of patients with ganglion cysts.57

In the ankle and foot area, injection therapy has been used to treat plantar fasciitis, tarsal tunnel syndrome, Achilles tendinopathy or bursitis, and interdigital neuroma (Morton’s neuroma). Most of the data about efficacy for these conditions are anecdotal and uncontrolled. Generally, the response in tarsal tunnel syndrome is temporary, whereas the response in Morton’s neuroma is more often prolonged. Reported response rates for plantar fasciitis vary, but one controlled trial showed a significant improvement at 1 month compared with placebo, whereas results were no different from placebo at 3 months.58,59 Studies of the efficacy of local corticosteroid injections for Achilles tendinopathy are inconclusive, and some have demonstrated an increased risk for Achilles tendon rupture.60

Local injections for neck pain and low back pain have been used for many years, with anecdotal reports of improvement, but controlled or prospective studies have shown variable results, depending on patient selection and methodology. Few controlled studies have assessed local trigger point or other soft tissue injections in the paracervical or paralumbar areas. Most studies of radiographically assisted facet joint injection of steroids in the lumbar or cervical areas show no difference compared with placebo, facet block, or local paraspinous injections.6163 Injections of the sacroiliac joint in patients with noninflammatory pain have shown a slight benefit from steroids compared with lidocaine alone.



All hydroxycorticosteroid preparations are effective for intra-articular and periarticular injections (Table 54-3). The originally injected hydroxycortisone acetate is still available, widely used, and inexpensive. Triamcinolone hexacetonide is one of the least soluble agents with the presumed most prolonged effect. Not all preparations are equivalent in efficacy or duration of effect, but few studies have been done to compare the efficacy of the various preparations. More recent reports have shown that methylprednisolone is superior to triamcinolone acetonide, that triamcinolone hexacetonide has a more prolonged response compared with triamcinolone acetonide, and both of these are more effective than hydrocortisone.64,65 Most clinicians have become familiar with certain preparations and have continued to use these with efficacy for years. Some clinicians prefer to inject combinations of short-acting and long-acting preparations. Steroid preparations are often mixed with local anesthetics, particularly for injecting small joints, tendon sheaths, and periarticular structures. Mixing with a local anesthetic reduces the local discomfort of injection into a confined space and dilutes the concentration of the locally injected steroid and reduces the risk of soft tissue atrophy. Guidelines for the dosage of steroid injected into given joints are based roughly on the size of the joint injected. Although no consensus exists regarding these amounts, most experts suggest injecting 1 mL of steroid preparation into large joints, with smaller amounts into smaller joints.

Table 54-3 Injectable Preparations for Intra-articular Injection

Corticosteroids Prednisone Equivalent (mg/mL)
Betamethasone sodium phosphate (6 mg/mL) 50
Dexamethasone sodium (4 mg/mL) 40
Dexamethasone acetate (8 mg/mL) 80
Hydrocortisone acetate (24 mg/mL) 5
Methylprednisolone acetate (40 mg/mL) 50
Prednisolone terbutate (20 mg/mL) 20
Triamcinolone acetonide (40 mg/mL) 50
Triamcinolone hexacetonide (20 mg/mL) 25
Hyaluronic Acid (Indicated in Osteoarthritis of Knee)
Hyaluronate Derivatives
Euflexxa: inject 20 mg (2 mL) once weekly for 3 wk  
Hyalgan: inject 20 mg (2 mL) once weekly for 5 wk; some may benefit with a total of 3 injections  
Orthovisc: inject 30 mg (2 mL) once weekly for 3-4 wk
Supartz: inject 25 mg (2.5 mL) once weekly for 5 wk  
Hylan Polymers
Synvisc: inject 16 mg (2 mL) once weekly for 3 wk (total of 3 injections)
Synvisc-One: inject 48 mg (6 mL) one time  

Other Injectable Products

Over the years, many other agents have been injected into joints including salicylates, phenylbutazone, gold, orgotein, progesterone, glycosaminoglycan polysulfate, and various antibiotics, but most have been abandoned because of lack of efficacy or local reactions. Various cytotoxic agents have been used sporadically or in small numbers of patients for intrasynovial tumors and refractory proliferative synovitis including nitrogen mustard, osmic acid, and methotrexate. Radioactive preparations (yttrium-90, colloidal 32P chromic phosphate, dysprosium-165-ferric hydroxide, rhenium-186) have been used in both inflammatory and noninflammatory arthritis in occasional reports, but the evidence for efficacy and safety of these substances in patients with arthritis is limited.66

Intra-articular hyaluronic acid preparations have been in use for many years in Europe and more recently in Canada and the United States. Several preparations of hyaluronic acid are approved for the treatment of osteoarthritis of the knee and seem to be superior to placebo injections in most, but not all, clinical trials, although no evidence for long-term efficacy or disease modification has been reported.37,38,6769 These preparations are usually given weekly in a series of three to five injections, and more recent studies have used single injections or higher molecular weight preparations.70 In direct comparisons with corticosteroids, hyaluronate is generally less effective in the first 4 weeks but more effective between 8 and 26 weeks after injections.71 Studies of hyaluronic acid preparations in osteoarthritic knees, hips, and shoulders have shown inconclusive or minimal levels of improvement.33,72 Hyaluronate injections may also be more effective than placebo in knees in rheumatoid arthritis and chronic painful rotator cuff disease, although these uses have not been as extensively studied.

Other injectable substances recently studied in small series for various conditions include botulinum toxin (for tennis elbow, painful shoulders, postoperative knee pain, and small joints in rheumatoid arthritis), EDTA for calcific tendinitis,73 and collagenase for Dupuytren’s contractures.74 In addition, platelet-rich plasma has been used for soft tissue injuries (e.g., Achilles tendinopathy, tennis elbow) with variable results in controlled studies.75


There are few contraindications to diagnostic arthrocentesis (Table 54-4). Established infection such as cellulitis in periarticular structures is generally considered to be an absolute contraindication to inserting a needle into a joint. If inflammation in an underlying joint or bursa is thought to be the cause of the appearance of infection, however, aspiration of the joint or bursa should be attempted. Septicemia carries the theoretic risk of introducing blood-borne bacteria into a joint, but such complications are not well documented and joints suspected of being infected should be aspirated regardless of the presence of septicemia. Arthrocentesis through an area of irregular or disrupted skin, as seen in psoriasis, should be avoided because of the increased numbers of colonizing bacteria in these areas. Caution should be exercised in patients with bleeding disorders or patients taking anticoagulants, owing to the theoretic risk of inducing hemarthrosis. The risk of significant hemarthrosis after arthrocentesis is low, however, even in patients on regular warfarin therapy with international normalized ratios of 4.5.76

Table 54-4 Contraindications to Arthrocentesis and Joint Injection

Contraindication Comment
Established infection in nearby structures (e.g., cellulitis, septic bursitis) Sometimes gout mimics cellulitis, creating a confusing picture
Septicemia (theoretic risk of introducing organism into joint) Need to tap suspected septic joints in septic patients
Disrupted skin barrier (e.g., psoriasis) Do not tap through lesions
Bleeding disorder (not absolute, but use more care) Risk of bleeding very low, even in patients taking warfarin
Septic joint Steroid injection contraindicated
Prior lack of response Relative contraindication
Difficult-to-access joint Relative contraindication without imaging aid


Iatrogenic infection is the most serious, but least common, complication of arthrocentesis and joint injection (Table 54-5). In Hollander’s large series,6 an incidence of infection of 0.005% was reported in a series of 400,000 injections. Gray and colleagues77 reported an incidence of 0.001% several years later. An infection rate of 1:2000 to 1:10,000 (0.01% to 0.05%) has been noted in patients with rheumatoid arthritis, occurring exclusively in debilitated patients on immunosuppressive therapy.78 A recent national database review in Iceland has estimated an infection rate of 0.037% from arthrocentesis.79 Few other prospective or systematic studies of infection after arthrocentesis have been published, but most reported anecdotal experience has noted a similar low incidence of this serious complication.80 An arthroscopic study showed that a small fragment of skin stained with a surgical marking pen could be identified within the joint space after most percutaneous insertions of a needle into the joint space, with identifications of bacterial nucleic acid by polymerase chain reaction in about one third of these.81 Considering the rarity of joint infection after arthrocentesis, these findings suggest that bacteria introduced at the time of arthrocentesis are either not viable or quickly cleared in almost all cases.

Table 54-5 Potential Complications of Arthrocentesis and Joint Injection

Complication Comment
Iatrogenic infection 0.01%-0.05%; may be higher in RA patients
Postinjection “flare” 1%-6%; lasting 48 hr; may be related to preparation
Local soft tissue May occur 1-6 mo later; pigment change
Local nerve damage In structures near prominent nerves (e.g., carpal tunnel syndrome)
Tendon rupture or weakening Case reports; animal studies show highest risk in Achilles tendon and plantar fascia
Systemic steroid absorption Inevitable; usually subclinical
Hypothalamic-pituitary suppression 2-7 days; changes in bone formation 14 days
Flushing; facial warmth; diaphoresis
Transient elevation of blood glucose; lymphopenia; eosinopenia
Avascular necrosis of bone Controversial; reported but usually explained by underlying disease or systemic steroids in same patient (ischemic necrosis)
Negative effects on cartilage Controversial
Found in animal models of normal cartilage but not in primates
Case reports in humans receiving multiple injections
Some animal models of arthritis are better with steroid injections
Large human observational studies have not documented more problems than expected (osteoarthritis, RA, juvenile RA)

RA, rheumatoid arthritis.

The most common complications of local steroid injections are related to local irritation of synovial and subcutaneous tissues and atrophy of soft tissues. Postinjection “flare” may develop in 1% to 6% of patients a few hours after injection and may last 48 hours, sometimes mimicking iatrogenic infection.6,8,82 These flares are reportedly more common with needle-shaped crystals and are believed to be similar to the acute arthritis related to other crystals phagocytosed by leukocytes, but they may also be caused by preservatives in some steroid suspensions.

Weakening of tendons and tendon rupture have also been reported as a result of locally injected steroids,83 emphasizing the importance of avoiding direct injection of steroids into the body of tendons. Most reports of tendon rupture have been anecdotal and described in patients involved in athletic activities or with rheumatoid arthritis. The risk of tendon rupture has not been adequately determined, but it seems to be quite low in the hands and wrists, where no ruptures were seen in a series of more than 200 injections84 and only 2 were seen in another series of 956 injections.8 Areas believed to be at highest risk for rupture include the Achilles tendon, bicipital tendon, and plantar fascia, where the risk of rupture has been estimated to be 10%.85

Systemic absorption occurs with locally injected depot corticosteroids. Since the earliest intra-articular injections of steroids, an anti-inflammatory effect has been shown not only in the injected joint but also in other joints in the same patient.4 Subsequent studies have documented decrease in plasma cortisol and suppression of the hypothalamic-pituitary axis lasting 2 to 7 days after a single injection. The degree and duration of adrenal suppression from a single intra-articular dose of depot steroid is less pronounced than that seen from an equal intramuscular dose.86 In a study of markers of bone turnover, a single injection of triamcinolone in knees of rheumatoid arthritis patients resulted in no change in bone resorption markers but yielded a drastic reduction in markers of bone formation within 1 day, which returned to normal levels in 14 days.87 A transient and variable effect on blood glucose levels after local steroid injection has also been observed in small studies, and unsurprisingly, this appears to be more pronounced in patients with diabetes.88

Some patients experience prominent erythema, warmth, and diaphoresis of the face and torso within minutes to hours after steroid injections.83 This reaction is most likely related to systemic absorption, but idiosyncratic reaction to preservatives in steroid preparations has also been implicated. Similarly, some patients may experience other typical metabolic effects of systemic steroids such as transient increases in blood glucose or decreases in peripheral blood eosinophil or lymphocyte counts.

Avascular necrosis of bone (ischemic necrosis) has long been considered a potential complication of intra-articular steroids, with a reported prevalence of this complication in injected joints ranging from less than 0.1% to 3%.6,18,23 Most studies have suggested, however, that the occurrence of this complication is related more to the severity of the associated disease or systemic steroid therapy and is unrelated to local injections.

The potential for negative effects of locally injected corticosteroids on cartilage metabolism has been a controversial area of study for several decades. Anecdotal reports of Charcot-like arthropathy attributed to intra-articular steroids first appeared in the late 1950s and 1960s, often occurring in patients having more than 10 (and sometimes hundreds) joint injections over many months or years. Several studies in the 1960s and 1970s showed that locally injected steroids caused destructive changes, catabolic effects, or both in normal animal cartilage.89,90 These included findings of decreased protein and matrix synthesis with degenerative cellular changes in chondrocytes, as well as fissures and decreased proteoglycan content in cartilage matrix. Similar studies done in primate joints failed to show any negative effects from intra-articular corticosteroids, however.91

Studies done in subsequent years have shown protective effects on cartilage lesions and reduction in osteophyte development in animal models of experimentally induced osteoarthritis, as well as associated reduction in metalloproteinase levels in cartilage and an increase in lubricating synovial surfactant.92,93 In humans with osteoarthritis, intra-articular steroids have been shown to decrease macrophage infiltration of the synovial lining, but no change was noted in metalloproteinase levels.94 Observations in humans treated with frequent corticosteroid injections have yielded conflicting information regarding changes in articular cartilage. More recent observations in patients with oligoarticular juvenile rheumatoid arthritis, mentioned previously, suggest that frequent steroid injections have the potential to help protect cartilage from the destructive process of the underlying disease process and are not associated with negative effects on articular cartilage.19,21,23 In addition, a study of patients with rheumatoid arthritis has shown no increase in the need for subsequent joint replacement surgery in the joints receiving four or more injections in a 1-year period.95

General Arthrocentesis Techniques


Most practitioners find that an arthrocentesis tray containing needed items allows more flexibility in preparing for aspirating or injecting joints or other articular structures. Syringes greater than 20 mL are not necessary for most procedures, but a swollen knee may occasionally contain 60 mL or more of fluid and it is reasonable to have at least one syringe of this size in a tray, with others available for rare patients with effusions greater than 100 to 200 mL in volume. Heparinized or citrated tubes to prevent coagulation of inflammatory fluids for accurate cell counts and crystal analysis, plain tubes for chemistry evaluations, and sterile tubes for transporting fluid to a microbiology laboratory for culture should be included. In joints being aspirated for the presence of bacteria and crystals, small amounts of fluid or debris may be present in the bore of the needle, even when no obvious fluid is obtained. In such situations, it is best to have clean microscope slides and coverslips available at the bedside for microscopic examination for cellularity, Gram stain, and crystals.

A hemostat is helpful for changing syringes after aspiration to inject corticosteroid. The size and length of needles used depend on the anticipated amount of fluid to be obtained from a joint and the size of the involved joint. A 20- to 22-gauge needle is usually sufficient to aspirate most detectable effusions, but large effusions with large amounts of debris, as seen in septic joints, may require larger-bore needles. In small joints, a 23- to 27-gauge needle may be used, particularly when no fluid appears to be present, and only therapeutic injection is being considered. A needle image inches in length is adequate for almost all procedures, but a 3-inch spinal needle may occasionally be necessary for a large knee or hip.

Site Preparation and Technique

Sterile technique designed to avoid the introduction of skin bacteria into the joint should be observed in all procedures, although precautions taken to avoid infection in clinical practice vary widely.80 After careful examination and identification of the specific point of aspiration, this point may be marked by the end of a ballpoint pen with the writing point retracted. The area should be carefully cleaned with one or two layers of iodine followed by alcohol. These precautions are sufficient to minimize the risk of infection, although a single “swipe” of isopropyl alcohol has been shown to provide a similar level of protection.96

A physician experienced in arthrocentesis may elect to not use topical anesthesia in many patients because the amount of pain is often no more than that experienced from phlebotomy. In an anxious patient or when small joints or joints with minimal fluid are being aspirated, topical anesthesia may be attained by the use of spray coolant (ethyl chloride) or an intradermal wheal and subcutaneous infiltration of lidocaine. Spray coolant may be applied after sterile preparation and has been shown to not contaminate the field.97 In pediatric patients, particularly when multiple joints are injected, sedation or general anesthesia may be required for safe and accurate injections.98 Nonsterile gloves should be worn by the operator to avoid contamination with the patient’s synovial fluid or blood. Drapes and sterile gloves are unnecessary, but the gloved hand should not touch the prepared site. A joint is usually entered at a 90-degree angle to the skin, slowly and evenly, and negative pressure should be applied to the syringe when the needle has been advanced image to 1 inch (in a large joint). If the needle’s course is obstructed by bone, the needle should be withdrawn slightly and redirected at a slightly different angle. If no fluid is obtained, the needle should be slowly advanced and negative pressure continued. If fluid flows initially and then stops, the needle may be advanced or retracted slightly or rotated in case it is blocked by an intra-articular structure or synovial tissue. After an adequate amount of fluid is obtained, a hemostat may be used to secure the needle, the syringe may be removed, and a new syringe with injectable steroid may be attached if injection is indicated. Overall accuracy of arthrocentesis has been estimated to be between 80% and 100% using anatomic landmarks in most joints.99 In “dry joints” a “backflow technique,” in which saline is injected and then aspirated back, may be used to help confirm correct needle positioning.100 A three-way stopcock is preferred by some practitioners when aspiration and injection are performed in the same setting.101

After injection, the needle and syringe should be removed and pressure applied over the site until a bandage is applied. When synovial fluid is to be examined for crystals, care should be taken to not replace the needle used for steroid injection on the syringe with synovial fluid because the contamination of fluid with steroid crystals would make accurate identification of urate and calcium pyrophosphate crystals more difficult. In some situations, positioning the barrel of the syringe and simultaneous aspiration may result in difficulty controlling needle position, and a newly available one-handed reciprocating syringe may allow for better operator control of the syringe in more difficult aspirations.102,103

< div class='tao-gold-member'>

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Jul 3, 2016 | Posted by in RHEUMATOLOGY | Comments Off on Arthrocentesis and Injection of Joints and Soft Tissue

Full access? Get Clinical Tree

Get Clinical Tree app for offline access