Article Text
PDF
Abstract
Background The use of diagnostic and interventional ultrasound has significantly increased over the past decade. A majority of the increased utilisation is by non-radiologists. In sports medicine, ultrasound is often used to guide interventions such as aspirations, diagnostic or therapeutic injections, tenotomies, releases and hydrodissections.
Objective Critically review the literature related to the accuracy, efficacy and cost-effectiveness of ultrasound-guided injections (USGIs) in major, intermediate and small joints; and soft tissues.
Design Systematic review of the literature.
Results USGIs are more accurate than landmark-guided injections (LMGIs; strength of recommendation taxonomy (SORT) Evidence Rating=A). USGIs are more efficacious than LMGIs (SORT Evidence Rating=B). USGIs are more cost-effective than LMGIs (SORT Evidence Rating=B). Ultrasound guidance is required to perform many new procedures (SORT Evidence Rating=C).
Conclusions The findings of this position statement indicate there is strong evidence that USGIs are more accurate than LMGI, moderate evidence that they are more efficacious and preliminary evidence that they are more cost-effective. Furthermore, ultrasound-guided (USG) is required to perform many new, advanced procedures and will likely enable the development of innovative USG surgical techniques in the future.
- Ultrasound
- Intervention
- Sports medicine
- Tendinopathy
- Osteoarthritis
Statistics from Altmetric.com
Background
The use of diagnostic and interventional musculoskeletal ultrasound (MSK US) in sports medicine has increased over the past several decades for a variety of reasons including decreased equipment costs, increased educational opportunities, expanded research, patient safety initiatives and technological advances leading to higher resolution images.1 Between 2000 and 2009, there was a 717% increase in the number of outpatient diagnostic MSK US studies, a majority of which were performed by non-radiologists.2 US can be used to diagnose disorders of bone, joints, tendons, muscles, ligaments, blood vessels and nerves as well as guide interventions such as aspirations, diagnostic or therapeutic injections, tenotomies, releases, hydrodissections and biopsies.3
As the utilisation of MSK US within sports medicine increases, it is important to critically review the existing literature and, based on the available evidence, make recommendations for its appropriate use. The purpose of this position statement is to evaluate the accuracy, efficacy and cost-effectiveness of US-guided injections (USGIs) in major, intermediate and small joints, and soft tissues, all of which are commonly performed in sports medicine. New procedures and future trends will also be briefly discussed.
Methods
Relevant English language articles through November 2013 were identified by searching Cochrane Database of Systematic Reviews and PubMed with the search terms injection, accuracy, efficacy, ultrasonography, fluoroscopy, joint and arthrography. The references of the articles were subsequently reviewed to identify additional articles not found in the original literature search. Articles that studied the accuracy, efficacy or cost-effectiveness of ultrasound-guided (USG) or landmark-guided injections (LMGIs) were included in the analysis for this position statement. Accuracy was defined as being able to place the injectate or needle tip in the intended structure. Studies that evaluated efficacy were defined as studies that evaluated a change in an outcome measure such as pain, range of motion, mobility, function or patient satisfaction following the procedure. Cost-effectiveness studies were defined as studies that evaluated the healthcare cost of the procedure relative to another treatment. The literature search was performed by a single researcher (MMH). An initial review of each study was subsequently performed by a separate researcher (JTF) and the level of evidence for each article was ranked according to the scale published by the Journal of Bone and Joint Surgery.4 For accuracy studies, the level of evidence was determined as follows: level 1—injections performed on live participants with accuracy confirmed using gold standard diagnostic imaging (ie, arthrogram for joints, MRI for soft tissues) or systematic review of level 1 studies; level 2—injections performed on live participants using non-gold standard imaging for accuracy confirmation, injections performed on cadaveric specimens with accuracy confirmed using gold standard diagnostic imaging or dissection, or systematic review of level 2 studies; level 3—injections performed on cadaveric specimens with accuracy confirmed using non-gold standard diagnostic imaging; level 4—injections performed on a small number (≤10) of live participants or cadaveric specimens, injections performed on live participants with accuracy confirmed by clinical outcome, or retrospective case series; level 5—case study or expert opinion. The literature was then distributed to the remaining authors for review and analysis. Disputes on classification were resolved via discussion and consensus. The literature was divided into the following categories for analysis: major joints, intermediate joints, small joints, multiple joints and soft tissues.
Results
The initial literature search identified 216 potential articles. Of these, 124 met the inclusion criteria for the position statement.
Major joints
Fifty-seven studies assessing injections in major joints were identified (see online supplementary appendix 1).5–61 A majority of the studies (49/57 (86%))6 ,7 ,9 ,11–13 ,15 ,17–25 ,27–30 ,32–51 ,53–61 evaluated injections in a single joint, whereas 14% (8/57)5 ,8 ,10 ,16 ,26 ,31 ,52 ,55 assessed injections in more than one joint. Thirty-five per cent (20/57) of the studies evaluated knee injections,8–10 ,13 ,15 ,16 ,19–23 ,26 ,31 ,32 ,36 ,37 ,48 ,52 ,56 ,57 46% (26/57) evaluated glenohumeral (GH) joint injections,5 ,7 ,8 ,10 ,11 ,15–17 ,24–26 ,28 ,29 ,31 ,38–40 ,42 ,43 ,46 ,47 ,49 ,52 ,54 ,60 ,61 21% (12/57) evaluated hip injections8 ,12 ,27 ,30 ,33 ,35 ,41 ,44 ,45 ,50 ,52 ,59 ,62 and 4% (2/57) evaluated sacroiliac (SI) joint injections.18 ,34 Four studies (7%) assessed injections in the ‘shoulder’, but did not specify which shoulder structure or joint they were injecting.6 ,53 ,55 ,58
The results of the studies investigating major joint injection accuracy are summarised in table 1. The level of evidence for a majority of the studies evaluating major joint USGI accuracy (15/23 (65%))5 ,8 ,9 ,15 ,16 ,18 ,20 ,21 ,23 ,27 ,30 ,32 ,34 ,36–39 ,41 ,42 ,44–46 ,49 or LMGI accuracy (28/28 (100%))5 ,7 ,9–13 ,17 ,19 ,21–26 ,28 ,31 ,32 ,36–38 ,40 ,43 ,47 ,48 ,50 ,60 ,61 were level 1 or 2. The mean accuracy of GH, hip and knee joint USGIs in studies with level 1 or 2 evidence ranged from 91% to 100%,5 ,8 ,9 ,15 ,18 ,21 ,23 ,32 ,36–39 ,42 ,46 ,49 whereas the mean accuracy of LMGIs were between 64% and 81%.5 ,7 ,9–13 ,17 ,19 ,21–26 ,28 ,31 ,32 ,36–38 ,40 ,43 ,47 ,48 ,50 ,60 ,61 These findings provide strong evidence that USGIs in the GH, hip and knee joints are more accurate than LMGIs.
Major joint injection accuracy
Only two studies investigated the accuracy of SI joint injections,18 ,34 and both studies only evaluated the accuracy of USGIs. While USG SI joint injections were 100% accurate in one of the studies,34 the other study reported an accuracy rate of only 40%.18 The discrepancy between the two studies may be due to a number of factors such as differences in accuracy assessment (colour Doppler-US vs MRI arthrogram), patient population, equipment variability, injector experience and injection technique. No studies were identified that evaluated the accuracy of landmark-guided (LMG) SI joint injections. Further studies are required to determine the accuracy of USG and LMG SI joint injections.
Nine studies with level 1 or 2 evidence investigated the efficacy of USGIs in major joints relative to LMGIs (see table 2).5 ,6 ,29 ,51 ,54–58 The joints evaluated in the studies included the GH joint (3 studies),29 ,51 ,54 shoulder (joint unspecified (3 studies))6 ,55 ,58 and knee joint (3 studies).5 ,56 ,57 Eighty-nine per cent (8/9) of these studies found that USGIs were more efficacious than LMGIs,5 ,29 ,51 ,54–58 while the remaining study found no difference in efficacy between the two injection techniques.6 A single study with level 1 evidence demonstrated no difference in efficacy between corticosteroid knee injections that were accurate versus those that were inaccurate.48 Based on the available research, in major joints, the majority of studies with level 1 or 2 evidence indicate that USGIs are more efficacious than LMGIs.
Major joint injection efficacy
Only two studies compared the cost-effectiveness of USGIs versus LMGIs (see table 3).56 ,57 Both of the studies were performed by the same group of researchers and evaluated the cost-effectiveness of USGIs and LMGIs in the knee. While both studies provided level 2 evidence suggesting that USGIs were more cost-effective than LMGIs, further research is required to corroborate their findings.
Major joint injection cost-effectiveness
In summary, USGIs in major joints other than the SI joint are more accurate than LMGIs. Further research is required to determine the accuracy of USG and LMG SI joint injections. The majority of evidence indicates USGIs in major joints are more efficacious than LMGIs in major joints. While the preliminary research suggests that USGIs are more cost-effective than LMGIs, further research is required before making a final determination on the cost-effectiveness of USGIs.
Intermediate joints
Twenty-three studies assessing injections into intermediate sized joints were identified (see online supplementary appendix 2).8 ,16 ,26 ,31 ,52 ,63–80 Seventy-four per cent (17/23) of the studies evaluated injections into a single joint8 ,63–65 ,67–70 ,72–80 and 26% (6/23) assessed injections into multiple joints.16 ,26 ,31 ,52 ,66 ,71 Injections into the following joints were evaluated: sternoclavicular (SC; 1/23 (4%)),79 acromioclavicular (AC; 7/23 (30%)),26 ,63 ,69 ,70 ,72 ,73 ,78 elbow (3/23 (13%)),16 ,26 ,31 wrist (4/23 (17%)),8 ,16 ,26 ,31 distal radioulnar (DRU; 1/23 (4%)),77 scaphotrapeziotrapezoidal (STT; 1/23 (4%)),74 proximal tibiofibular (TF; 1/23 (4%)),76 tibiotalar (TT; 7/23 (30%)),16 ,26 ,31 ,65 ,66 ,71 ,80 subtalar (ST; 5/23 (22%)),66–68 ,71 ,75 and midfoot (1/23 (4%)).64
Twenty-one of the 23 studies (91%) assessed intermediate joint injection accuracy (see table 4).8 ,16 ,26 ,31 ,52 ,63 ,65–72 ,74–80 Their findings are summarised in table 4. Similar to the injection accuracy studies in major joints, a majority (20/21 (95%)) of the intermediate joint injection accuracy studies provided either level 1 or level 2 evidence.8 ,26 ,31 ,52 ,63 ,65–72 ,74–80 In the studies with level 1 or 2 evidence, the mean accuracy of USGIs into intermediate joints ranged from 95% to 100%.8 ,52 ,63 ,66 ,70 ,71 ,74–77 ,80 The mean accuracy of LMGIs into intermediate joints with level 1 or 2 evidence was between 0% and 92%.26 ,31 ,52 ,63 ,65–70 ,72 ,74 ,76 ,78–80 The accuracy of LMGIs varied widely by joint and approach.
Intermediate joint injection accuracy
The only study that evaluated injection accuracy into the SC joint used a LMG approach, and reported a mean accuracy of 78%.79 Since no USGI studies into the SC joint have been performed, a comparison of SC joint injection accuracy between the two techniques cannot be made.
Two level 2 studies evaluated USGI accuracy into the AC joint, and reported mean accuracy of 95%.63 ,70 Five level 2 studies63 ,69 ,70 ,72 ,78 evaluated the accuracy of LMG AC joint injections and reported a mean accuracy of 52%. In addition to accuracy, the results presented by Sabeti-Aschraf72 looked at USGI and LMGI accuracy of three subgroups: physician specialist, physician non-specialist and student. As expected, the student's LMGI accuracy was the lowest (60%) and the physician specialist's LMGI accuracy was the highest (80%). When the same providers used USG, accuracy improved to 90–100% with the students being the highest of the three subgroups. Based on the available evidence, USGIs into the AC joint are significantly more accurate than LMGIs.
Two level 1 studies evaluated LMGI accuracy into the elbow joint.26 ,31 The mean accuracy of these studies was 97%. The only study evaluating the accuracy of USGIs into the elbow joint provided level 4 evidence that elbow joint USGI accuracy was 100%. The current research suggests that elbow joint LMGIs are quite accurate and, although preliminary findings imply that elbow joint USGIs are also accurate, further research is required to corroborate this data.
The accuracy of injections into three different sites about the wrist has been studied. The first is the DRU joint (DRUJ). A single level 2 study reported the accuracy of USGIs into the DRUJ to be 100%.77 No DRUJ LMGI accuracy studies were identified. A single level 1 study demonstrated 100% accuracy of wrist joint USGIs.8 The mean accuracy of wrist joint LMGIs reported by two level 2 studies was 74%.26 ,31 A single level 2 study demonstrated the accuracy of STT joint injections using USG to be 100%, while LMGI accuracy was 80%.74 Therefore, initial findings indicate USGI accuracy into the distal RU, wrist and STT joints is 100% accurate, but further research is required to confirm these conclusions. The current evidence suggests LMGIs into the wrist and STT joints are less accurate than USGIs (74% and 80%, respectively, vs 100%), and no research is available regarding the accuracy of distal RU joint LMGIs. However, due to the paucity of research on injections in the wrist region, further research is required before definitive conclusions can be drawn.
The accuracy of injections into three intermediate sized, lower extremity joints (proximal TF, TT and ST joints) has been studied. A level 2 study reported proximal TF joint USGIs to be 100% accurate, while LMGIs into the same joint were 58% accurate.76 TT joint USGIs were found to be 100% accurate in three level 2 studies.66 ,71 ,80 The mean TT joint LMGI accuracy was 64% in two level 1 studies26 ,31 and 87% in three level 2 studies.65 ,66 ,80 The mean ST joint USGI accuracy of three level 2 studies was 97%,66 ,71 ,75 while three level 2 studies reported the accuracy of LMGI to be 89%.66–68 These findings suggest that proximal TF, TT and ST joint USGIs are highly accurate, while LMGIs into the same regions have variable accuracy, with the highest level of accuracy found in the ST joint (89%).
Finally, one level 2 study evaluated the accuracy of USGIs and LMGIs into multiple joints (elbow, wrist and TT joints).52 Balint et al reported 100% accuracy of USGIs into the elbow and TT joints, while the mean accuracy of LMGIs into the elbow, wrist and TT joints was only 29%. However, the conclusions of this study are significantly limited based on the small number of injections performed.
Four studies evaluated the efficacy of intermediate joint USGIs versus LMGIs (see table 5).16 ,26 ,64 ,73 One was a level 2 study,73 another was a level 3 study26 and the remaining two were level 4 studies.16 ,64 Sabeti-Aschraf et al73 found no difference in efficacy between AC joint USGIs and LMGIs. Jones et al26 found no difference in efficacy between accurate and inaccurate injections into the AC, elbow, wrist and ankle joints, but the conclusions of this study are limited due to the study design. Both level 4 studies demonstrated that USGIs were efficacious into intermediate joints.16 ,64
Intermediate joint injection efficacy
No studies evaluated the cost-effectiveness of USG versus LMG intermediate joint injections.
In summary, USGIs into a majority of intermediate joints are more accurate than LMGIs, although LMGIs into the elbow and ST joints were relatively accurate (mean accuracy of 97% and 89%, respectively). However, most joints only had one or two studies evaluating injection accuracy. Therefore, further USG and LMG intermediate joint injection accuracy studies are necessary to make definitive conclusions regarding intermediate joint injection accuracy. Despite the difference in accuracy between USG and LMG intermediate joint injections, the only study that evaluated the difference in efficacy between the two injection techniques did not find a difference.73 Interestingly, the joint evaluated in this study (AC joint) was one of the joints with a fairly large difference in accuracy between USGIs and LMGIs (95% vs 52%). Since they did not evaluate the accuracy of their injections, it is difficult to determine whether the lack of difference in efficacy between the two techniques is because they had similar accuracy rates between the two techniques, or because efficacy is not related to accuracy in this particular joint. Owing to the paucity of research, a definite conclusion regarding whether or not USG improves the efficacy of intermediate joint injections cannot be made.
Small joints
Nine studies assessing injections in small joints were identified (see online supplementary appendix 3).16 ,26 ,31 ,52 ,66 ,71 ,81–83 A small majority of these studies (5/9 (56%))31 ,66 ,71 ,82 ,83 evaluated injections into a single type of small joint (eg, metacarpophalangeal (MCP) joint), while the remainder (4/9 (44%))16 ,26 ,52 ,81 evaluated injections into multiple small joints. Sixty-seven per cent (6/9)16 ,26 ,31 ,52 ,81 ,82 of the studies assessed small joint injections in the hands and 56% (5/9)16 ,52 ,66 ,71 ,83 evaluated small foot joint injections. Of those studies assessing hand procedures, three studies (50%) included the carpometacarpal (CMC) joint,26 ,52 ,82 two (33%) the PIP joints52 ,81 and one (17%)26 the distal interphalangeal (IP) joints. Among the studies of foot procedures, four (60%)16 ,52 ,71 ,83 were directed at the metatarsophalangeal (MTP) joints and one (20%)66 the tarsometatarsal (TMT) joints.
The results of the studies investigating small joint accuracy are summarised in table 6. The majority (5/8 (63%)) of small joint injection accuracy studies provided level 1 or 2 evidence.31 ,66 ,71 ,82 ,83 The remaining studies provided level 3 or 5 evidence.26 ,52 ,81 In the hand, a single level 2 study reported the mean USGI accuracy of the CMC joint to be 94%.82 There were no level 1 or 2 studies for LMGI accuracy of the CMC joint. A single level 3 study compared the accuracy of USG and LMG CMC joint injections and found the mean accuracy of USGI to be 100% and of LMGI to be 0%.52 No study was identified that addressed the accuracy of USG MCP joint injections, but a single level 1 study reported the mean accuracy of LMGI to be 97%.31 No level 1 or 2 studies evaluated the accuracy of IP joint injections. One level 3 study compared the accuracy of USG versus LMG IP joint injections and found the mean accuracy of USGI to be 100%, while the accuracy of LMGI was 0%.52 Another level 3 study reported the accuracy of USG MCP and PIP joint injections to be 96% and LMGI to be 59%.81
Small joint injection accuracy
Regarding small joint injections in the feet, a single level 2 study compared the accuracy of USG and LMG TMT joint injections and found the USGIs to be more accurate (64% accurate) than LMGIs (25% accurate).66 Three studies (two with level 2 evidence71 ,83 and one with level 3 evidence52) found 100% accuracy for USGI of the MTP joints with one of the three52 noting poor accuracy (0% accurate) with LMGI.
Only a single, level 4 study addressed the efficacy of USGI of the small joints (see table 7).16 This case series demonstrated that USGI of the MCP and MTP joints were efficacious, but the strength of their findings was limited due to a lack of a comparison group.16 No studies were identified that compared the cost-effectiveness of USGI versus LMGI of the small joints. Thus, it is unclear from the available literature whether the superior accuracy suggested by the available studies translates into improved outcomes or cost savings.
Small joint injection efficacy
In summary, current research suggests that USGIs in small joints are more accurate than LMGIs. However, due to the paucity of high-quality research evaluating small joint injection accuracy, further research is required to confirm these initial findings prior to drawing final conclusions. There is insufficient evidence at this time to determine whether USG small joint injections are more efficacious or cost-effective than LMGI.
Soft tissues
Forty-nine studies assessing injections into soft tissues were identified (see online supplementary appendix 4).17 ,51 ,52 ,54 ,69 ,71 ,80 ,84–125 Most studies evaluated injections into a single structure (42/49 (86%)),17 ,51 ,54 ,69 ,80 ,84–87 ,90–95 ,97–101 ,103–113 ,115–125 but seven studies (14%) investigated injections into more than one structure.52 ,71 ,88 ,89 ,96 ,102 ,114 In decreasing frequency, studies evaluated injections into bursae (19/49 (39%)),17 ,51 ,52 ,54 ,69 ,87 ,91–93 ,95 ,97 ,100 ,101 ,103 ,105 ,110 ,115 ,121 ,123 tendon sheaths (9/49 (18%)),71 ,89 ,102 ,106 ,108 ,111–113 ,116 tendons or fascia (8/49 {(16%)),96 ,102 ,107 ,112 ,119 ,120 ,124 ,125 perineural regions (6/49 (12%)),85 ,88 ,94 ,104 ,109 ,122 muscles (5/49 (10%)),86 ,97 ,114 ,117 ,118 cysts (2/49 (4%)),84 ,90 peritendinous regions (2/49 (4%)),71 ,102 wounds (1/49 (2%))52 and periarticular spaces (1/49 (2%)).80
Soft tissue injection accuracy studies are summarised in table 8. Four level 1 or 2 studies evaluating the accuracy of tendon sheath or peritendinous injections were identified.71 ,99 ,113 ,116 Multiple regions were evaluated including the Achilles peritendinous region,71 and the tendon sheaths of the long head biceps,99 first dorsal wrist compartment, flexor hallucis longus,71 tibialis posterior,71 popliteus116 and peroneal (fibularis) tendons.113 Although the criteria used to define ‘accurate injections’ were different in the various studies, the mean reported accuracy of USGIs into tendon sheaths or peritendinous regions ranged from 87% to 100%, while the mean accuracy of LMGIs ranged from 27% to 60%. Thus, there is strong evidence that USG tendon sheath or peritendinous injections are more accurate than LMGIs.
Soft tissue injection accuracy
Ten level 1 or 2 studies examined the accuracy of subacromial-subdeltoid (SA-SD) bursa injections.17 ,69 ,91 ,93 ,95 ,101 ,105 ,110 ,115 ,123 As with peritendinous injections, the definition of an ‘accurate injection’ was not uniform among the studies. Accuracy rates for LMG SA-SD bursa injections ranged from 24% to 100%, while USGI accuracy ranged from 65% to 100%. Although USG SA-SD bursa injections were more consistently accurate than LMGIs, due to the highly variable results reported across different studies, a definite conclusion regarding whether or not USG SA-SD bursa injections are more accurate than LMGIs cannot be made at this time. Further research is required to clarify this question.
A single level 2 study evaluated the accuracy of LMGI versus USGI into the pes anserinus bursa.98 The accuracy rate for LMG pes anserinus bursa injections was 17%, while USGI accuracy was 92%. These preliminary findings suggest that USG pes anserinus bursa injections are more accurate than LMGIs.
One level 2 study compared the accuracy of USG piriformis injections to fluoroscopically guided injections.97 US guidance provided accurate injections in 95% of cases, while fluoroscopically guided injections were accurate only 30% of the time. Furthermore, one of the fluoroscopically guided injections placed the injectate into the sciatic nerve. Another level 2 study reported the accuracy of USG obturator internus injections to be 100%.118 Although preliminary, these findings suggest US guidance enables accurate injections into the deep gluteal musculature, is more accurate than fluoroscopically guided injections into this region and minimises the potential for complications associated with inadvertent needle placement into adjacent neurological structures.
A level 1 study evaluated the accuracy of placing the needle tip of a compartment pressure monitor into the deep and superficial posterior leg compartments using landmark or US guidance in cadavers.114 The accuracy was similar between the two techniques. This was likely due to the relatively superficial location and large size the two posterior leg compartments. Therefore, based on the current evidence, USG is not recommended for routine compartment pressure testing of the posterior leg compartments.
Two level 2 studies evaluated the accuracy of USGIs into Morton's neuromas.94 ,104 Both reported 100% accuracy. No studies were identified that evaluated the accuracy of LMG Morton's neuroma injections. Based on the available evidence, USG Morton's neuroma injections are highly accurate and the accuracy of LMG Morton's neuroma injections is unknown.
The final soft tissue injection accuracy study was a level 2 study that evaluated the accuracy of LMG sinus tarsi injections versus USGIs.80 Wisniewski et al reported the accuracy of USG sinus tarsi injections to be 90%. LMGIs were only 35% accurate. These findings suggest that USG sinus tarsi injections are more accurate than LMGIs.
Regarding efficacy, only one study was identified with level 1 or 2 evidence that directly compared LMGIs to USGIs for the treatment of a tendon disorder (see table 9).106 Kume et al demonstrated significantly more pain reduction from USGIs than LMGIs in patients with septation between the extensor pollicis brevis and abductor pollicis longus tendons in the first dorsal compartment. Septation is present in the first dorsal compartment in greater than 50% of patients.111 Although further studies are needed, USGIs for the treatment of de Quervain's tenosynovitis may be superior to LMGIs, particularly in the setting of a septated first dorsal compartment.
Soft tissue injection efficacy
Two level 2 studies compared the efficacy of USG plantar fascia injections versus LMGIs.119 ,125 Neither of the studies found any difference in efficacy between USG plantar fascia injections and LMGIs, although one of the studies reported less recurrent pain following USGIs.119 In addition, one of the studies evaluated the efficacy of scintigraphically guided plantar fascia injections compared with USGIs and LMGIs.125 No difference in outcomes was found between the three groups. Interestingly, ‘scintigraphic guidance’ was actually an unguided injection since the injector performed a LMGI in the region where the scintigram was positive. There is currently insufficient evidence to support routine US guidance for plantar fascia injections. Further studies are needed to determine whether USG plantar fascia injections reduce recurrence rates, which may decrease the costs associated with treating this condition. Finally, research is also required to determine whether US guidance reduces complications associated with plantar fascia injections (eg, plantar fascia rupture, calcaneal fat pad atrophy).
Five level 2 studies evaluated the efficacy of USG SA-SD bursa injections versus LMGIs.51 ,54 ,87 ,103 ,121 All five studies demonstrated better outcomes following USG SA-SD bursa injections compared with LMGIs. Three level 2 studies assessed the efficacy of accurate versus inaccurate SA-SD bursa injections.91 ,93 ,105 Two of the studies concluded there was no difference in efficacy between accurate and inaccurate injections,91 ,105 and one reported that accurate injections are more efficacious than inaccurate injections.93 A single level 2 study demonstrated more pain relief following USG SA-SD bursa local anaesthetic injections than LMGI, suggesting USG SA-SD bursa injections may provide more diagnostic information regarding the aetiology of shoulder pain than LMGIs.100 A final level 2 study demonstrated more improvement in a majority of outcome measures following USG SA-SD bursa injections than oral steroids for shoulder pain.92 Therefore, current studies indicate USG SA-SD bursa injections are more efficacious than LMGIs or oral steroids for shoulder pain. Furthermore, USGIs provide more diagnostic information regarding the aetiology of shoulder pain than LMGIs.
Three level 2 studies compared the efficacy of USG carpal tunnel injections to LMGIs.85 ,109 ,122 All three studies reported that USG carpal tunnel injections were less painful and more efficacious than LMGIs. Furthermore, one of the studies performed a cost analysis and concluded that USG carpal tunnel injections were also more cost-effective than LMGIs (see table 10).85 However, the cost analysis only included those who responded to the injection. When all patients were included in the cost analysis (responders and non-responders), the cost was higher for USGIs than for LMGIs when the procedure was performed in a physician's office, and was equivalent when performed in a hospital-based setting. The findings of these studies provide strong evidence that USG carpal tunnel injections are more efficacious than LMGIs. However, further research is required to determine if USG carpal tunnel injections are more cost-effective than LMGIs.
Soft tissue injection cost effectiveness
In summary, USGIs into tendon sheaths, peritendinous regions, deep gluteal muscles (eg, piriformis and obturator internus), the pes anserinus bursa and sinus tarsi are all more accurate than LMGIs. USG Morton's neuroma injections are highly accurate, but the accuracy of LMGIs into Morton's neuromas is unknown at this time. Although USG SA-SD bursa injections appear to be more accurate than LMGIs, the wide range of reported accuracy limits the ability to draw a definitive conclusion at this time. USG SA-SD bursa, carpal tunnel and first dorsal wrist compartment injections are more efficacious than LMGIs. USG plantar fascia injections appear to have equivalent outcomes to LMGIs. Finally, further research is required to determine if USGIs into soft tissues are more cost-effective than LMGIs.
Multiple joints
Three studies were identified that evaluated joint injections in multiple locations (see online supplementary appendix 5).126–128 None of the three studies specified which joints were assessed. The accuracy, efficacy and cost-effectiveness data from these studies are summarised in tables 11⇓–13. The first study evaluated the efficacy and cost-effectiveness of USGIs versus LMGIs into joints with inflammatory arthritis.127 This level 2 study found that USGIs into joints with inflammatory arthritis produced less procedural pain, more pain relief, more responders and less non-responders to the injection, and was less expensive than LMGIs. In a study with level 1 evidence, Cunnington et al126 determined that the mean accuracy of USGIs into joints with inflammatory arthritis was 83% accurate, while LMGIs were only 66% accurate. Their study also provided level 2 evidence that USGIs into joints with inflammatory arthritis resulted in more clinical improvement and pain reduction at 6 weeks follow-up than those who received a LMGI. The final multiple joint injection study performed by Sibbit et al128 provided level 2 evidence that participants with painful joints who received USGIs experienced less procedural pain and more pain relief than those who received LMGIs. Moreover, when compared with LMGIs, USGIs resulted in a larger number of responders, less non-responders, and an improved ability to detect and aspirate joint effusions.
Multijoint injection accuracy
Multijoint injection efficacy
Multijoint injection cost-effectiveness
In summary, these findings suggest that USGIs into inflamed or painful joints are more accurate, less painful, more efficacious and less expensive than LMGIs. However, further research is required to confirm these findings due to the limited number of studies.
New procedures and future trends
As the field of MSK US has continued to mature, practitioners from multiple disciplines have capitalised on US's powerful combination of high (submillimeter) resolution and real-time imaging capability to expand the applications of interventional MSK US in clinical practice. These applications can be considered in three broad categories, or generations.
First generation techniques apply US guidance to improve the accuracy of established procedures such as joint injections, peritendinous injections and perineural injections, and are the focus of the current position statement. The use of first generation techniques has continued to expand as additional therapeutic and regenerative agents have been introduced into clinical practice, including but not limited to dextrose, autologous blood and platelet rich plasma.129–139 This trend will continue as practitioners utilise US guidance as the primary deployment mechanism to deliver an increasing repertoire of drugs, cell-based therapeutic-regenerative agents and tissue scaffolds to soft tissues and accessible joint regions.62 ,140–143
Second generation techniques have predominately emerged during the past decade and can be generally considered to be advanced procedures performed with commonly available needles. However, in contradistinction to first generation techniques, most of the second generation techniques were developed primarily as a result of the availability of US guidance. Common examples include needle tenotomy/fasciotomy for chronic tendinosis/fasciosis, fenestration of the transverse carpal ligament to treat carpal tunnel syndrome, neovessel ablation via sclerosing agent injection or mechanical disruption to treat chronic tendinosis, needle release of the A1 pulley for trigger finger, needle aponeurotomy for Dupuytren's contracture, and hydrodissection to treat peripheral neuritis due to mild compression or adhesions.62 ,140 ,141 ,143–154 Prior to the widespread adoption of US guidance, these procedures either did not exist, or were performed relatively rarely due to the inability to directly visualise target tissues and subsequent safety concerns. Currently, many of these procedures are being increasingly utilised on a regular basis in diverse clinical practices. Percutaneous US-guided fenestration and aspiration (ie, barbotage) of calcific tendinosis can also be considered to be a second generation procedure. Although originally described as a fluoroscopic procedure, the role of fluoroscopy has largely been supplanted by US guidance due to US's excellent safety profile and clinical efficacy.155–158
Third generation techniques are perhaps the most exciting for the field, and are characterised by the use of pre-existing, specialised surgical tools or specially designed devices to perform a specific US-guided procedure. Many of these techniques duplicate well-accepted surgical procedures using percutaneous US guidance to improve safety and reduce morbidity. Recently described techniques include A1 pulley release using hook knives, carpal tunnel release using hook knives, arthroscopic equipment or specially designed devices; and tenotomy/fasciotomy using specialised devices that not only cut but also debride damaged tissue.159–167 The integration of these techniques into clinical practice represents a major advancement in the field of MSK medicine. In the near future, it is likely that additional USG surgical procedures will be adopted with advanced US imaging techniques and/or specialised equipment.
In summary, the current trend towards expanded applications of interventional MSK US can be expected to continue for decades, driven by advances in US technology, practitioner expertise with US guidance and the development of specialised tools. Many traditional surgical procedures will become office-based, lower cost procedures performed by skilled practitioners, and some will be combined with precise delivery of therapeutic-regenerative agents.
Discussion
The purpose of this position statement was to determine the accuracy, efficacy and cost-effectiveness of USGIs in joints and soft tissues. A brief discussion of new USG procedures and future trends was also conducted. During the following discussion, the American Medical Society for Sports Medicine (AMSSM) position on each topic will be stated, and the strength of the evidence associated with the position will be graded using the following strength of recommendation taxonomy (SORT):
-
Consistent, good-quality evidence;
-
Inconsistent or limited-quality evidence;
-
Consensus, disease-oriented evidence, usual practice, expert opinion or case series.
Accuracy
AMSSM Position: USGIs are more accurate than LMGIs (SORT Evidence Rating=A)
A majority of the relevant research investigated USGI accuracy. There is evidence that USGIs into large, intermediate and small joints; tendon sheaths, peritendinous regions, deep gluteal muscles, pes anserinus bursa, sinus tarsi and inflamed joints are more accurate than LMGIs. The preponderance of studies evaluated the accuracy of large joint injections followed by intermediate joints with the minority of studies evaluating the accuracy of small joint injections. Owing to the limited number of small and intermediate joint injection accuracy studies, further research in these areas is warranted.
Preliminary research suggests that USGIs into Morton's neuromas are highly accurate, but no LMGI accuracy studies have been performed so a comparison between the two techniques cannot be made. Similarly, no LMGI accuracy studies have been performed in the SI joint and the two USG SI joint injection studies that have been published reported conflicting accuracy rates. Therefore, further research is required to determine whether USG SI joint and Morton's neuroma injections are more accurate than LMGIs.
The soft tissue structure with the most injection accuracy studies was the SA-SD bursa. Although a majority of research suggested that USG SA-SD bursa injections are more accurate than LMGIs, the reported accuracy rates for both USGIs and LMGIs were highly variable. This may have been due to several factors. First, USGIs are only accurate if the injector can correctly identify the target and guide the needle into the target. Therefore, the variability of the USGI accuracy results suggests that the injectors in some USG SA-SD bursa injection studies were either unable to accurately identify the SA-SD bursa or correctly guide the needle into the target. Since the injector's ability to correctly identify the SA-SD bursa was not assessed, nor was their ability to guide a needle into a specific target, the influence of the injector's technical abilities on the studies outcome is unknown. The technique by which accuracy is confirmed may also have influenced the study outcomes. For instance, in the study by Mathews and Glousman,110 20 cadaveric shoulders were injected with radiocontrast into the SA bursa using two different approaches, and the accuracy of the injections was initially determined by fluoroscopy to be 90%. However, after dissecting the shoulders, the actual accuracy rate was determined to be 60%. This demonstrates that imaging modalities cannot always be relied on to provide correct information regarding injection accuracy, particularly into soft tissues. The heterogeneity of accuracy confirmation techniques (CT, CT arthrography, MRI, MR arthrography, standard radiographic arthrography, intraoperative confirmation, cadaveric dissection) employed by different researchers contributes to the difficulty of interpreting the injection accuracy literature. Further research in which the injector's technical abilities are confirmed and the correct imaging technique is used to grade accuracy are required to definitively answer the question of whether or not USG SA-SD bursa injections are more accurate than LMGIs.
Efficacy
AMSSM Position: USGIs are more efficacious than LMGIs (SORT Evidence Rating=B)
There is evidence that USGIs are more efficacious than LMGIs in large joints, inflamed joints, SA-SD bursa, carpal tunnel and first dorsal wrist compartment tendon sheath. Only one study evaluated the efficacy of USG intermediate joint injections (AC joint) relative to LMGIs and found no difference in efficacy between the two techniques, but the study's design limits the strength of their conclusions. No studies have been performed comparing the efficacy of USG small joint injections to LMGIs. Therefore, although a majority of studies suggest that USGIs are more efficacious than LMGIs, further research is required to fully answer this question.
There are some difficulties with performing efficacy research that warrant mention. The most commonly injected substance to treat MSK conditions is corticosteroids. There is limited evidence that the systemic effects of corticosteroids provide similar therapeutic benefits to localised injections.92 In the study by Ekberg et al,92 a corticosteroid injection in the gluteal region was compared with an USG SA-SD bursa injection for patients with rotator cuff disease. While their conclusions need to be interpreted with caution due to significant study limitations (eg, heterogeneity of shoulder pathology in the study participants, lack of control group, soft tissue corticosteroid injections in both groups which may result in larger systemic effects than intra-articular injections, etc), both groups showed similar improvements in their primary outcome measures, although there were some secondary outcome measures that were better in the USGI group than the gluteal (systemic) injection group. Therefore, it is possible that the systemic effects of corticosteroids may make it difficult to detect a difference in efficacy between an accurately and inaccurately placed corticosteroid injection. Despite this possibility, it is important to remember that several studies have been able to demonstrate greater efficacy with accurately placed corticosteroids than inaccurately placed corticosteroids. This may be due to the type of pathology that is being treated. Specifically, although corticosteroids have been demonstrated to provide short-term therapeutic benefits for arthritis,168 it can be argued that corticosteroid injections may not be an effective treatment for some conditions such as rotator cuff tendinopathy.169 So, the issue of injection accuracy and efficacy may be irrelevant if the injected agent (eg, corticosteroids) is inappropriate for the pathology being treated. Certainly one could postulate that injectable therapeutic agents that do not have demonstrable systemic therapeutic benefits (eg, viscosupplements, platelet rich plasma) would be ineffective if placed in the wrong region. Therefore, therapeutic benefit would be dependent on correct injectate placement for these compounds. However, further research is required to determine if this hypothesis is correct.
While the difference in efficacy between USGIs and LMGIs is important, since it has been established that LMGIs are less accurate than USGIs, it is also important to consider the non-therapeutic ramifications of inaccurate injectate placement. If an injectate is misplaced, it may lead to complications such as skin depigmentation, subcutaneous fat atrophy, tendon rupture, neurovascular injury, increased procedural and postprocedural pain or intra-arterial injection.99 ,108 In addition, correct injectate placement can provide useful diagnostic information regarding the location of a pain generator. All of these factors must be taken into consideration when choosing which injection technique to employ.
Cost-effectiveness
AMSSM Position: USGIs are more cost-effective than LMGIs (SORT Evidence Rating=B)
The area with the least research is cost-effectiveness. Only four studies were identified that compared the cost-effectiveness of USGIs to LMGIs. The preliminary findings of these studies suggest that USGIs are more cost-effective than LMGIs for large joints, inflamed joints and carpal tunnel syndrome since more people responded to the USGIs, their improvement was greater and lasted longer than those who received LMGIs, and they utilised healthcare services less often following USGIs than LMGIs. However, due to the limited number of studies, additional well-designed studies are required to determine if USGIs are more cost-effective than LMGIs.
New procedures and future trends
AMSSM Position: USG is required to perform many new procedures (SORT Evidence Rating=C)
Finally, the scope of USG procedures in sports medicine continues to evolve with the introduction of second generation (eg, tenotomies, transverse carpal ligament fenestrations, peripheral nerve hydrodissections) and third generation (eg, percutaneous A1 pulley releases with a surgical hook knife) procedures. Direct visualisation of the target structure, relevant surrounding structures, and guidance of the procedural device is required for the performance of these procedures. Although the need for radiological guidance (eg, USG) is inherent to the performance of these procedures, research will be needed to determine the efficacy, safety profile and cost-effectiveness of these new procedures.
Conclusions
The use of diagnostic and interventional US has significantly increased over the past decade. A majority of the increased utilisation is by non-radiologists. In sports medicine, US is often used to guide interventions such as aspirations, diagnostic or therapeutic injections, tenotomies, releases and hydrodissections, and is rapidly becoming part of the standard practice of sports medicine. The findings of this position statement indicate there is strong evidence that USGIs are more accurate than LMGIs, moderate evidence that they are more efficacious, and preliminary evidence that they are more cost-effective. Furthermore, USG is required to perform many new, advanced procedures and will likely enable the development of innovative USG surgical techniques in the future.
What are the new findings?
-
The findings of this position statement indicate that there is strong evidence that ultrasound-guided injections (USGIs) are more accurate than landmark-guided injections (LMGIs). There is also moderate evidence indicating USGIs are more efficacious and cost-effective than LMGIs. Finally, in the author's opinion, many new procedures require guidance in order to perform the procedure safely and effectively.
How might it impact on clinical practice in the near future?
-
Based on the findings of this position statement, sports medicine physicians should consider using ultrasound guidance when performing soft tissue and joint injections, or when performing procedures that should utilise guidance such as percutaneous needle tenotomies or lavage and aspiration of calcific deposits.
Acknowledgments
The authors would like to acknowledge Sasha Rupp for her contributions to the creation of this position statement.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online appendix 1
- Data supplement 2 - Online appendix 2
- Data supplement 3 - Online appendix 3
- Data supplement 4 - Online appendix 4
- Data supplement 5 - Online appendix 5
Footnotes
-
Author note This position statement will also be published in Clinical Journal of Sports Medicine, and PM&R Journal.
-
Collaborators Sasha Rupp.
-
Contributors JTF, MMH, EA, DB, ALC, WD and JS all were involved in the planning of the position stand. MMH performed the initial literature search with the assistance of a reference librarian. JTF followed by the remaining authors reviewed the results of the initial literature search and provided additional references. JTF ranked the level of evidence of each study and created the appendices and tables. This information was subsequently reviewed and revised by the rest of the authors. MMH wrote the initial draft of the section on soft tissue procedures. JTF wrote the initial draft of the major joint and multiple joint injection sections. ALC wrote the initial draft of the small joint injection section. EA wrote the initial draft of the intermediate joint injection section. JS wrote the initial draft of the new procedures and future trends section. All of the sections were then reviewed and revised first by JTF, then the entire group of authors. The authors then discussed the results of each section, determined the appropriate recommendation for injection accuracy, efficacy, cost-effectiveness, and new procedures and ranked the level of evidence behind these recommendations using strength of recommendation taxonomy. The first draft of the recommendations and conclusions sections were then written by JTF and reviewed by the remaining authors. The final draft was then reviewed by the entire authorship to identify any remaining issues, and after all recommendations had been incorporated into the document and every revision had been made, the document was submitted.
-
Competing interests None.
-
Provenance and peer review Not commissioned; externally peer reviewed.