Physical therapy intervention versus corticosteroid injection for lateral elbow tendinopathy. Does slow and steady win the race? – A systematic review

Abstract

Background

Lateral elbow tendinopathy is one of the most common chronic and degenerative diseases which significantly affects quality of life and the activities of daily living of a person. The following is a systematic review reporting a comparison between physical therapy intervention and corticosteroid injection for the treatment of lateral elbow tendinopathy.

Method

PubMed, Web of Science, and Embase were searched using headings related to treatment options for Lateral elbow tendinopathy. The following keywords were used: lateral epicondylitis, physical therapy, and corticosteroid injection.

Result

We descriptively analyzed and reviewed a total of 12 studies including a total of 1253 patients for lateral elbow tendinopathy. The physical therapy intervention included interventions like electrotherapy, manual therapy, and exercise. The studies included had an overall low to unknown risk of bias.

Conclusion

Our review suggests corticosteroid injection provides beneficial short-term effects and physical therapy interventions provide intermediate to long-term effects, less additional treatment and low recurrence rate in patients with lateral elbow tendinopathy. Although high-quality randomized control trials are required in order to have a better understanding of both intervention types.

Introduction

Tennis elbow coined by Runge, presently also pronounced as lateral elbow tendinopathy is one the most common chronic and degenerative diseases which significantly affects life quality and activities of daily living of a person.^1,2 Up to 3% of the population is affected by lateral elbow tendinopathy, with 15% being heavy industrial workers. Most beginners in tennis, practicing the one-handed backhand type of shot are the victims of it. ³ The pathophysiology of lateral elbow tendinopathy consists of the integration of inflammation, microscopic tearing, and degenerative process. ⁴ Such microscopic tears can lead to macro trauma at the proximal insertion of extensor carpi radialis as per Nirschl and Pettrone. ⁵ According to Cyriax, ⁶ the disease patho-physiology involved microscopic or macroscopic tears of the extensor origin. Clinical features include elbow pain over the lateral aspect reproduced with palpation, which may radiate in the forearm, and painful wrist extension with a load. ⁷ The classic complaints by the patient will be a pain, along with strength deficits in wrist extension and gripping activities. ⁸ Various physical therapeutic, medical, and surgical treatment options are established as per literature with advancements in recent years. Some treatment options are nonsteroidal anti-inflammatories (NSAIDs), corticosteroid injections, splints, electrotherapeutic modalities, and exercise therapy. ⁹ Due to a deficiency of high-quality randomized control trials regarding comparative analysis of management protocols, a gold standard treatment protocol is yet to be determined.

On one aspect, corticosteroid injection efficacy has attracted many clinicians for the treatment of lateral elbow tendinopathy. ¹⁰ Corticosteroid injection yield both anti-inflammatory and analgesic mechanisms through synovial blood flow reduction, leukocyte genomic suppression, production of cytokines and protease, and collagen synthesis.^11–14 The other aspect of this, physical therapy interventions (PTIs) such as electrotherapeutic modalities and exercise therapy are one of the conventional forms of treatment options across the globe. The efficacy of PTIs and rehabilitation science in patients with lateral elbow tendinopathy was reported by various systematic reviews integrating various interventions, for example, ultrasound therapy, exercise, and manual therapy.^15–18

Therefore, the escalated use of corticosteroid injection treatment in recent years compared to the PTIs in many clinical trials has led to dissimilar conclusions. A systematic review and meta-analysis in the year 2009 comparing solely the corticosteroid injection over one arm and PTI on the other concluded corticosteroid therapy was a better treatment option for short-term results but PTI had better long-term results. ¹⁹ Considering recent literature, the introduction of various randomized clinical trials comparing both interventions separately forms the basis of this review.

Material and methods

Eligibility criteria:

Inclusion criteria

1. Randomized controlled trials included patients with lateral elbow tendinopathy.

2. Studies that compared corticosteroid injection with the group of either one or more PTIs like electrotherapy or manual therapy (MT) or exercise.

Exclusion criteria

1. Studies that used interventions other than corticosteroid injection or PTI.

2. Studies that compared corticosteroids with other interventions for the same condition.

3. Studies with languages other than English.

Search strategy and information sources

We followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-analyses) reporting guidelines from Cochrane Collaboration. Using headings related to treatment options for lateral elbow tendinopathy. PubMed, Web of Science, and Embase were searched till June 2022. The following keywords were used: lateral epicondylitis, physical therapy, and corticosteroid injection. We manually screened references from reviews. The search was conducted only in the English language by SS and PB with the final decision being taken by NV in case of differences of opinion. A detailed search strategy is provided in Supplemental File Table S1.

Study selection

Those studies compared PTI with corticosteroid injection as well as studies the effects of corticosteroid injection combined with PTI to sham injection or PTI alone for lateral elbow tendinopathy.

Data extraction

Two reviewers (SS) and (PB) independently extracted data in terms of participant demographics, study characteristics, applied interventions, outcomes, and conclusion. Figure 1 depicts the extracted results of the search. The eligible studies were concluded for quality assessment as per PRISMA guidelines.

Figure 1.

Flowchart of the search strategy.

Study risk of bias assessment

The risk of bias assessment was done separately by two reviewers (SS and PB) in accordance with the Cochrane handbook's quality assessment. In case of difference of opinion final decision was taken by a third author (AKP). The studies were assessed for selection bias, performance bias, detection bias, attrition bias, and reporting bias. The presence of publication bias could not be eliminated via a funnel plot.

Outcomes

A variety of outcomes was used in the included full-text articles. This included pain score, DASH (Disabilities of the arm, shoulder, and hand questionnaire), Quick DASH, pinch strength, extensor weight strength, pain pressure threshold (PPT), Nirschl Pain Phase Scale of Athletic Overuse Injuries (NPPS), Patient Related Elbow Evaluation (PREET), Patient-Rated Tennis Elbow Evaluation (PRTEE), PRFEQ, PGART, Euro-QoL5 ED, ultrasonography, and extensor thickness measurement (Supplemental File Table S2).

Results

Study selection

Out of 942 records screened, 628 were obtained after eliminating duplicates. Further title and abstract screening eliminated 604 studies. A total of 24 studies were found eligible for thorough full-text assessment. Out of which 12, studies were eliminated because they failed to fulfill the inclusion criteria (Supplementary File Table S2).

Study characteristics and risk of bias

Supplementary File Table S2 shows the cases detailed characteristics of the included studies. We reviewed a total of 12 studies including a total of 1253 patients for lateral elbow tendinopathy. Among the 12 studies, two studies used ultrasound therapy (UST),^17,20 three studies used deep transverse friction massage (DTFM),^17,20,21 two studies used Shockwave therapy,^20,22 one study used pulsed electromagnetic field therapy (PEMF), ²³ two studies used kinesiotaping (KT),^24,25 one study used dry needling (DN), ²⁶ three studies used mobilization^21,27,28 and five studies incorporated exercise program^27–31 as PTI. For corticosteroid injection eight studies^{21,22,24, 27–30,31} used triamcinolone derivatives and four studies^20,23,25,26 used prednisolone. Table 1 displays the included study characteristics.

Table 1.

Included Studies Charactertistics.

Author, Year	No. of partcipants	Mean Age	Duration of symptoms	Physical therapy intervention	Corticosteroid intervention	Outcomes Assessed	Follow up period	Inclusion criteria	Exclusion criteria
1. VERHAAR 1995 (21)	106	43	33 weeks	Physio- therapy consisted of deep transverse friction massage over the exten- sor origin and Mills’ manipulations. A total of 12 treatments was given over four weeks, followed by another two weeks of restriction of all painful activities	1 ml of triam- cinolone acetate suspension 1% diluted with 1 ml of lido- caine 1% into the tendinous origin of the extensor digitorum and extensor carpi radialis brevis muscles. The patients were seen two and four weeks after the start of the treatment and a second or third injection was given if there were persistent symptoms at the two- or four-week follow- up	Severity of pain Occurrence of pain Subjective loss of grip strength Resumption of labour Mean grip strength Patients’ assessment and result rating	-0 week - 6 weeks -52 weeks	pain on the lateral side of the elbow, tenderness over the forearm extensor origin, and pain on the lateral epicondyle during resisted dorsiflexion of the wrist with the elbow in full extension	previous operation on the lateral side of the elbow, or had arthritis or allied conditions, neurological disorders of the painful extremity, more than three local corticosteroid injections during the six months before consultation, and if the same elbow had been treated before by Cyriax’s methods
2. Smidt 2002 (29)	185	W and See- 46(42-54) CSI Group-47(41-54) PT group- 48(41-52)	more than 6 weeks	Physiotherapy consisted of nine treatments of pulsed ultrasound, deep friction massage, and an exercise programme over 6 weeks. Pulsed ultrasound (20% duty cycle) was given with an intensity of 2 W/cm2 for 7·5 minutes per session for 9 sessions. exercise treatment consisted of progressive, slow, repetitive wrist and forearm stretching, muscle conditioning, and occupational exercises, intensified in four steps.	local infiltration of 1 mL triamcinolone acetonide (10 mg/mL) and 1 mL lidocaine 2%. Every tender spot was identified with the needle and injected until the patient was free of pain during resisted dorsiflexion	Main complaint, Pain during day, Inconvinience, Severity of elbow complaints, Elbow disability Pain-free grip strength	-6 weeks -12 weeks -26 weeks -52 weeks	pain at the lateral side of the elbow, increasing with pressure on the lateral epicondyle and with resisted dorsiflexion of the wrist; age 18–70 years; ability to complete questionnaires in Dutch; and informed consent	treatment of elbow complaints with physiotherapy or injections during the preceding 6 months; bilateral elbow symptoms; duration less than 6 weeks; presence of signs and symptoms suggestive of another cause of elbow pain—eg. cervical radiculopathy; congenital or acquired deformities of the elbow; surgery of the elbow; dislocation, tendon ruptures, or fractures in the area in the preceding 12 months; systemic musculoskeletal or neurological disorders; and contraindications for corticosteroids
3. Crowther 2002 (22)	93	49 years (27 to 69)	16weeks or more	three sessions of ESWT at weekly intervals. Treatment was provided by an ultrasonographer from United Medical Systems using the portable StorzMinilith SL1 lithotripter. A total of 2000 shock waves (maximum 0.1 MJ/mm2) was administered at each session with inline ultrasound guidance	injection of 20 mg of triamcinolone made up to 1.5 ml with 1% lignocaine using an aseptic technique, into the point of maximal tenderness at the extensor origin of the lateral epicondyle of the humerus	VAS	-6 week -3 month	over 18 years old with a classic history of tennis elbow for longer than four months and no surgical intervention or injection in the previous year. Positive clinical findings of tenderness over the lateral epicondyle of the humerus and reproducible pain with resisted finger and wrist extension	evidence of dysfunction of the shoulder, neck or thorax, local arthritis, generalised polyarthritis, generalised neurological abnormality, nerve entrapment in the upper limb, pregnancy, infection, tumour, a clotting disorder, were on anticoagulant therapy or had a cardiac pacemaker
4. Uzunca 2006 (23)	75	PEMF Group-46.76(6.24) SHAM PEMF Group- 51.47(8.23) CSI Group-47.80(6.05)	more than 6w	Active PEMF therapy was performed in Group 1 by a magnetotherapy device (BTL-09, manufactured by BTL Benesov, Czech Repub- lic). The injured elbow of each patient was put in the middle portion of a big circle solenoid applicator in prone position. The dose and application time were selected according to the recommendations of the manufacturer. The total dose applied was 6 mT/session. This dose was completed by applying the PEMF in a frequency of 25 Hz and a frequency of 4.6 Hz, consecutively. A therapy session lasted for 30 min and 15 sessions were performed during 3 weeks (five sessions a week for 3 weeks).-	Local corticosteroid injection was administered to the most painful area with pressure around lateral epicondyle Injected material consisted 1 cc of methylprednisolone acetate (40 mg) and 1 cc of prilocaine hydrochloride (20 mg).	Rest pain (VAS) Activity pain (VAS) Night pain (VAS) Pain during resisted wrist dorsiflexion (VAS) Pain during resisted forearm supination (VAS)aAlgometric pain threshold (kg/cm2)	- 3 week -3 month	lateral elbow and forearm pain that lasted for more than 6 weeks, tenderness in the origin of the extensor carpi radialis brevis muscle and increased tenderness of dorsiflexion of the wrist against resistance and of forearm supination	accompanying painful conditions, which may confuse the clinical picture such as upper extremity fracture, inflammatory arthritic conditions, carpal tunnel syndrome, thoracic outlet syndrome, cervical radiculopathy, and tendon ruptures; accompanying medial epicondylitis; contraindications for PEMF such as tuberculosis, pregnancy, cardiac pacemaker, and malignancy; and contraindications for corticosteroid injection
5. Bisset 2006 (27)	198	47.6	6 weeks	programme of elbow manipulation and therapeutic exercise.	corticosteroid injections with the full amount of a local injection, consisting of a 1 ml quantity of 1% lidocaine with 10 mg of triamcinolone acetonide in 1 ml, delivered to painful elbow points	-Global improvement Pain-free grip force - Assessor’s rating of severity.	-3 week -6 week -12 week -26 week - 52 week	Pain over the lateral elbow that increased on palpation of the lateral epicondyle, gripping, resisted wrist, or second or third finger extension, Age 18-65 years with pain of at least six weeks duration.	Any treatment of the elbow pain by a healthcare practitioner within the preceding six months, Bilateral elbow symptoms, Cervical radiculopathy, any other elbow joint pathology, peripheral nerve involvement, previous surgery to the elbow history of dislocation, fracture of the elbow, or tendon ruptures, systemic or neurological disorders;  shoulder, wrist, and hand pathology; contraindications to corticosteroids
6. Tonks 2007 (30)	48	PT Group- 43.8(7.5) CSI Group-48.2(6.5)	24 weeks	Physiotherapy consisted of the exercise programme devised by Pienimaki et al. and comprised progressive slow, repetitive wrist and forearm stretching and muscle conditioning, intensified into four steps.	single injection of 10 mg of triamcinolone acetonide and of 2% lignocaine hydrochloride made up to a volume of 1 ml injected into the symptomatically tender region of the common extensor origin	PFGS (kg) EWS (kg) PRFEQ – pain PRFEQ – function PRFEQ – total	-7 week	patients with symptoms of tennis elbow who had not had treatment for this complaint in the preceding 6 months, pain reproduced on palpation of the common extensor origin, pain reproduced on resisted extension of the wrist with the elbow extended and age over 18 years	Trauma to the affected elbow in the preceding 6 weeks Patients with a past history of elbow instability Previous elbow surgery Bilateral symptoms Any other pathology involving the affected upper limb Coexisting cervical spine pathology Physiotherapy or steroid injection for the presenting condition within the previous 6 months Patients already on oral/systemic steroids  Patients with contraindications to injection therapy: Patients with bleeding diatheses or on anticoagulant therapy Local or systemic infection Past history of hypersensitivity to local anaesthetics  Poorly controlled diabetics and other immunosuppressed individuals Pregnant or breast feeding patients Uncooperative patients with an underlying psychiatric diagnosis Prosthetic elbow joint
7. Gündüz 2012. (20)	59	PT group- 43.6(9.1) CSI group- 45.7(10.2) ESWT group- 44.9(9.9)	less than 12 weeks	physical treatment modalities including hot pack (15 min), ultrasound therapy (1 W/cm2, 5 min) and friction massage (5 min) for ten sessions. ESWT group (n 0 20) was treated with ESWT (pressure 1.4 bar, frequency 4.0 Hz, number 500) for ten sessions (1 day interval between the sessions).	received a single injection of 20 mg methylprednisolone acetate and 1 ml prilocaine	-VAS -Grip strength -Pinch strength	-1 month -3 month -6 month	pain on the lateral side of the elbow that was severe enough to interfere with their daily living activities (for less than 3 months), tenderness over the lateral epicondyle compared with that of the normal elbow, and pain during extension of wrist and fingers against resistance	previous treatment for ipsilateral LE, other elbow pathologies, cervical vertebrae/upper limb disorders, previous elbow surgery, joint limitations due to a previous radius/ulna fracture, osteoporosis, malignancy, he- mophilia, neurological deficit(s) in the ipsilateral upper limb, and cognitive dysfunction.
8. Coombes 2013. (28)	165	49.7(8.1)	6 weeks	manual therapy and exercises from the protocol and progressed the program based on the patients’ capabilities to allow for optimal exercise volume and load setting without exacerbating pain. The specific elbow manipulation (mobiliza- tion with movement) techniques were applied in combination with gripping	Patients received a single injection of either placebo (0.5 mL of 0.9% isotonic saline) or corticosteroid and local anesthetic medication (10 mg/mL of triamcinolone acetonide in a 1 mL injection plus 1 mL of 1% lignocaine)	-PRTEE -VAS -EQ-5ED	- 4 week - 26 week -52 week	pain over the lateral humeral epicondyle with pain severity of greater than 30 mm on a 100-mm visual analog scale (VAS), provoked by at least 2 of the following: gripping, palpation, resisted wrist or middle finger extension, or stretching of forearm extensor muscles with reduced pain-free grip	Exclusion criteria were receipt of injection (during preceding 6 months); receipt of a course of physiotherapy (during preceding 3 months); concomitant neck or other arm pain necessitating treatment or preventing participation in usual work or recreational activities (during preceding 6 months); symp toms suggesting radicular, neurological, or systemic arthritic conditions; pregnancy; breastfeeding; or contrain- dication to injection
9. Küçükşen 2013 (31)	82	MET group-46.17(7.56) CSI group- 43.78(9.16)	12 weeksor more	MET was applied (isometric contraction approximately 75% of maximal) against resistance for a period of 5 seconds	1mL of triamcinolone acetonide (40mg/mL) plus 1mL of 1% lidocaine (10mg/mL) were injected deep into the subcutaneous tissues and muscles, 1cm distal to the lateral epicondyle	PFGS VAS DASH	6w 26 52	tenderness on or near the lateral epicondyle and if the pain was elicited with at least 2 of 3 pain provocation tests (gripping, resisted wrist extension performed with the elbow in extension, and stretching of the forearm extensor muscles).unilateral elbow pain for >3 months, pain severity 50mm on a 100-mm visual analog scale (VAS), age between 18 and 70 years, and willingness to comply with treatment and follow-up assessments.	treatment of elbow complaints with surgical intervention; physiotherapy or CSIs in the past 6 months; bilateral elbow symptoms; duration of <3 months; severe neck or shoulder problems likely to cause or maintain the elbow complaints, as determined by the investigator; posterior interosseous nerve compression; congenital or acquired deformities of the elbow; systemic musculoskeletal or neurologic disorders; age <18 years; and contraindications for corticosteroids (pregnancy or breast-feeding).
10. Koçak 2019. (25)	84	43.06 ± 11.19 yrs	2-12 weeks	elastic therapeutic tape was based on the recommended application techniques by Kase et al	The SI was performed while the patient's forearm was resting in pronation and 45-degree flexion on a firm surface. After ensuring standard aseptic conditions, the mixture of 20 mg of methylprednisolone acetate (0.5 mL) and 0.5 mL of prilocaine at 2% was injected into the subcutaneous tissues and muscles 1 cm distal to the lateral epicondyle in a fan-shaped manner using a 22-G 30-mm needle.	-VAS at rest -VAS during movement -Pain-free grip strength -PPT -qDASH -PGART	-3 week -12 week	Patients with pain at the lateral side of the elbow, whose pain increased with pressure on the lateral epicondyle and opposed wrist extension, opposed middle-finger extension, or passive stretch of the wrist extensors, with a pain duration between 2 and 12 weeks, whose pain severity was equal to 5 and higher on a numeric scale where 0 indicated no pain and 10 indicated the most severe pain, who were aged 18–70 yrs, and those with sufficient cognitive ability to complete questionnaires and sign the informed consent	Patients who received treatment for LE; with problems in both elbows; with a history of inflammatory arthritis, elbow surgery, trauma to the elbow, tendon rupture of the muscles located around the elbow, humerus, radius or ulna fracture or joint dislocation, oral or systemic steroid use, malignancy; limb joint motion limitation; congenital or acquired elbow joint deformity; cervical vertebrae or other upper limb problem; root pathology of cervical origin; systemic musculoskeletal or neurologic dis- ease; cooperative difficulty due to cognitive dysfunction; and conditions for which local steroids were contraindicated (septic arthritis, sepsis, tuberculosis, unknown caused fever, history of allergy against local anesthetics or steroids, monoarthritis with unknown cause, neutropenia, thrombocytopenia, anticoagulant use, coagulation disorders)
11. Uygur 2020. (26)	108	DN Group-47.5(7.3) CSI Group-48.1(10.3)	more than 12 weeks	dry needling with fifteen 25-mm stainless steel needles inserted at the lateral epicondyle region and throughout the course of the extensor carpi radialis brevis tendon (Fig. 1, a). The needles were placed down to the bone.	all CS (2 mL methylprednisolone acetate, DepoMedrol", 40 mg/mL) injections (single doses) were performed by the second author who employed ‘‘peppering’’ using a 22-G needle (Fig. 1, b). The periosteum was pricked 20-30 times without withdrawing the needle	PRTEE	-20 days -6m	Patients with LE pain for more than months whose pain did not improve with first line of treatment including NSAIDS and Forearm brace for weeks	recalcitrant cases (greater than 18 months of pain prior to treatment),prior elbow surgery or history of elbow trauma, invasive treatment for LE within 3 months before study enrollment, and patients with inflammatory arthritis or uncontrolled diabetes (patients whose serum glucose levels are under control with oral antidiabetic drugs and not higher than 150 mg/mL)
12. Erpala 2021 (24)	50	46.8(8.4)	6 weeks	The Kinesio tape was applied to the patients in the KT group by two certified authors (FE and UB) using muscle and area correction techniques as described by Kase et al.	10 mg triamcinolone (20 mg/mL 0.5 cc) without dilution by any agent was used. Under sterile conditions, after proper staining, a single injection with a 22-G 30 mm needle was applied by the same authors (FE and UB) to the most painful point in the origin of the ECR on the lateral epicondyle.	-Nirsch LE scoring scale -QDASH -PREE-T -VAS -VAS EXT -Hand grip strength	- 2 week - 4 week	Patients aged 18–70 years who tested positively in at least one of the Mills, Maudsley or Cozen’s tests and experienced symptomatic pain for at least 6 weeks lateral to the elbow	Patient’s with cervical radiculopathy, Neuropathy, Previous diagnosis of rheumatological disease, Autoimmune disease, Diabetic neuropathy, Surgical histories, Open wounds and scars, Local or systemic infections, Pregnancy, History of allergic reactions to tape or corticosteroid injections or elbow joint athrosis Patients who had CSI in the elbow area in the area in the last 6 months

Risk of bias assessment

We discovered that the majority of studies had a low risk of bias in most domains. Although all studies claimed to follow randomization, one study didn’t clarify methods used for allocation concealment ²⁶ and another showed high risk due to ineffective concealment. ²⁴ Six studies did not clarify blinding processes.^{20–22,24,25,30} Figures 2 and 3 represent the risk of bias for overall and individual studies.

Figure 2.

Risk of bias assessment (overall).

Figure 3.

Risk of bias assessment (individual study).

Results of individual studies

We came across three RCTs that compared corticosteroid injection with PTI as well as wait-and-see policy. These studies concluded PTI is superior to the rest of the two alternatives in the long run. As per one study ²⁷ with a low risk of bias, the majority of participants experienced relief with injections [51/65 participants (78%)] compared with a wait and see [16/60 participants (27%)]. The injection was superior to PTI in the majority of outcome measures. However, in the long term, the participants of the injection group reported significantly worse outcomes in comparison with the PTI group (0.3, 0.1–0.5; number needed to treat = 4) and wait-and-see (0.3, 0.04–0.4; 4).

The second study, ²⁹ however, reported corticosteroid injections to be superior to all other therapy options for all outcome measures. Success rates were 92% (57 patients) as compared to 47% (30 patients) for PTI and 32% (19 patients) for the wait-and-see policy. The injection group showed the highest recurrence rate. The long-term success rates were 69% (43 patients) for injections, 91% (58 patients) for PTI, and 83% (49 patients) for a wait-and-see policy. The study used specific methods for randomization, followed concealed allocation, and did not report incomplete outcomes.

On the contrary, the third study ³⁰ reported corticosteroids to provide better improvement in pain-free grip strength (improvement of 6.24 points, 95% CI: 0.15–12.33), PRFEQ (improvement of 2.73 points, 95% CI: 1.44–4.03) as well as in extensor weight strength (improvement of 1.849 points, 95% CI: 0.77–2.93). It thus stated corticosteroid injection solely as the primary line of treatment as per the conclusion. The study however did not clarify blinding methods and completion outcome data was not clear.

There was also a study comparing the effects of physical therapy intervention and corticosteroid injections with a combination of both. As per the RCT ²⁸ which had a low risk of bias Corticosteroid injection group had lower complete recovery at 1 year versus placebo injection (83% vs. 96%, respectively; relative risk [RR]: 0.86 [99% CI: 0.75–0.99]; p = 0.01) and greater recurrence at 1 year (54% vs. 12%; RR: 0.23 [99% CI: 0.10–0.51]; p = 0.001), whereas PTI and no PTI groups did not differ much on improvement (91% vs. 88%, respectively; RR: 1.04 [99% CI: 0.90–1.19]; p = 0.56) or recurrence (29% vs. 38%; RR: 1.31 [99% CI: 0.73–2.35]; p = 0.25). No differences were found among patients receiving the corticosteroid injection along with PTI versus corticosteroid alone (68% vs. 71%, respectively; RR: 0.95 [99% CI: 0.65–1.38]; p = 0.57).

We also found two studies reporting the results of KT versus corticosteroid injections in lateral elbow tendinopathy. One prospective study with a follow-up of 4 weeks ²⁴ reported a slight decline in the functional scores of the rest and corticosteroid groups compared to KT. The KT group had a superior effect over the rest and the medication group. Even though allocation concealment was improper, and the blinding methods were not clear, the study concluded KT be the sole effective treatment in the final week even though no significant differences were observed between the KT, the RMG, and the CSI groups in the fourth week.

In another single-blinded trial with a 12-week follow-up, ²⁵ corticosteroid injection was given to one group, another group received KT, and the third group received both interventions. The co-administration of corticosteroid injection along with KT was to be more effective compared to each treatment alone. The study did not specify the blinding of outcomes.

Another study ³¹ compared the Muscle Energy Technique to Corticosteroid Injection. MET was reported to have superior scores for PFGS and the VAS in the long term, even though improvements in the CSI were significantly better initially. After 52 weeks, the MET group showed a significantly higher PFGS score (75.08 ± 26.19 vs. 62.24 ± 21.83; p = 0.007) along with a significantly lower mean VAS score (3.28 ± 2.86 vs. 4.95 ± 2.36; p = 0.001) compared to CSI group.

We also found a single prospective RCT ²⁶ which compared the effect of corticosteroid injection with DN. The authors reported a decrease in PRTEE scores in both groups however, the DN group (60.9 ± 11.8 to 15.6 ± 7.7) showed a greater difference compared to the CS group (58.6 ± 5.1 decreased to 36 ± 14.7). At 6 months of follow-up, it was concluded that DN-treated patients showed better improvement (p < 0.01). The authors did not mention concealed allocation and outcome data seemed incomplete.

Another clinical trial ²³ compared the effects of PEMF, sham PEMF, and a corticosteroid combined with an anesthetic agent injection in terms of VAS and pain threshold. Initially, most of the pain parameters showed significant improvement in the PEMF group when compared with the sham PEMF group at the end of 3 months all the pain parameters showed improvement after therapy in all three groups however the improvement was greater in the corticosteroid group and least in the sham PEMF Group. The trial however showed the unknown risk of attrition bias.

A prospective, randomized trial ²¹ observed the effects of corticosteroid injections versus physical therapy intervention as described by Cyriax. The authors failed to specify methods of concealment, blinding was not mentioned and reporting bias was unknown. Even though only half of the patients, received the initial treatment, 20% had combined therapy and 30% had surgery, the study reported a greater increase in grip strength in the physical therapy group (−2.3 ± 10.6 at 6 weeks to 11.0 ± 13.8 at 52 weeks) compared to corticosteroid (10.7 ± 14.9 at 6 weeks to 14.6 ± 13.1 at 52 weeks).

We also found two studies reporting the effectiveness of corticosteroid injection and extracorporeal shock-wave therapy. Even with known mention of blinding, one study ²² reported corticosteroids (mean pain score from 66 to 21) provide better relief in pain initially compared to shockwave (mean pain score from 61 to 35). At the end of 3 months corticosteroid group (84%) had a better reduction in pain than the shockwave group (60%). The other study reported a decrease in VAS scores in all the groups up to 6 months of treatment however there was no change in Pinch strength and ultrasonographical findings. Another RCT ²⁰ reported a decrease in VAS scores of the patients however grip strength showed a significant increase in the CSI group initially and later in the PTI group and the ESWT. No group showed any change in pinch strength and ultrasonographical findings, the authors concluded that PTIs, corticosteroid injection, and ESWT have an effect on grip strength along with pain in the initial period of lateral elbow tendinopathy treatment. The improvement in grip strength lasts longer with ESWT. The authors however failed to mention the blinding of outcome assessment.

Taking into consideration the additional treatment given to patients in the group, Bisset et al. ²⁷ reported that 32/65 patients of the CSI group needed additional treatment [NSAID (n = 20), elbow support or brace (n = 10), complementary medicine (n = 13)], 13/63 patients in PTI group [analgesic or NSAID (n = 9)].

Coombes et al. ²⁸ reported that 17/43 CSI patients and 7/41 PTI patients needed additional treatment of NSAIDs. In the study by Smidt et al., ²⁹ 39/62 CSI patients [more CSI (n = 21), PTI (n = 13)] and 52/64 patients [more PTI (n = 42), CSI (n = 4)] required additional treatment. Other studies didn’t report the information for the same.

Discussion

Overview

Our review accounted for the quality assessment of 12 randomized control trials after a thorough literature search. The use of both physical therapy intervention (PTI) and corticosteroid intervention is gaining acceptance as a conservative form of treatment for lateral elbow tendinopathy. Both interventions helped in managing the symptoms of the patients which varied with short and long-term follow-up. A previous review was conducted in the year 2010 under the same criteria. The gist of the study projected corticosteroid injection is beneficial for short-term benefits and PTI is beneficial for intermediate to long-term benefits. Due to the recent addition of various RCTs to the literature regarding the comparison of corticosteroid injection and PTI, our review takes into account a much wider picture providing a head-to-head comparison of the two treatment options. Also, our study highlights the heterogeneity in the methodology of the included studies.

Physical therapy intervention

Since physical therapy and therapeutic modalities are quite common in the modern practice of medicine. The physical therapy interventions used in the included studies varied under three subgroups: electrotherapy, MT, and exercise regimen.

Ultrasound therapy

The three electrotherapy interventions used were therapeutic UST, extracorporeal shockwave therapy, and PEMF therapy. Only five studies reported the use of electrotherapy which varied with the dosage, frequency, and duration of a similar type of modality. The dosage and technique of UST in study 6 was continuous mode, UST (1 W/cm², 5 min) for 10 sessions whereas in study 7 was pulsed mode UST (20% duty cycle) with 2 W/cm² intensity, for 7.5 min per session for nine sessions. Although there is a shred of weak evidence to defend the use of UST for various musculoskeletal disorders. ³² Also, pulsed UST was not considered a much better option than placebo in the case of chronic lateral epicondylitis. ³³

Extracorporeal shockwave therapy

In the case of ESWT, study 6 used ESWT with 1.4 bar pressure, 4.0 Hz frequency, and 500 shots for 10 sessions (single day interval between the sessions). On the contrary, in study 11, 2000 shock waves (maximum 0.1 MJ/mm²) were administered along with ultrasound guidance at each session for a total of three sessions (one session per week). Also, a Yuan meta-analysis done in 2019 revealed ESWT superior to corticosteroids in lateral elbow tendinopathy management. Therefore, ESWT has shown promising results in comparison to corticosteroid injection. Only one study reported the use of active PEMF therapy in included studies. A study done in 1985 was done to evaluate the difference between active or inactive PEMF which reported no significant difference for lateral elbow tendinopathy. ³⁴ High-quality RCTs are required to deduce the practicality of PEMF for LE.

Manual therapy

In MT, elbow mobilization,^1,2 DTFM,^6,7 MET, ⁵ Cyriax-type and physical therapy technique ¹⁰ which consisted of mill's manipulation + DTMF, KT,^3,4 and DN. ⁸ The type, dosage, and duration of these interventions also varied in each included study for the same treatment technique. A meta-analysis done in 2020 reported DN intervention as low to moderate evidence of short-term benefits on pain, disability, pressure pain sensitivity, and strength for lateral elbow tendinopathy. ³⁵ An RCT reported DTFM with lidocaine injection better than cortisone injection. ³⁶

Exercise therapy

A home exercise therapy program was followed in the study by Bissett et al., ²⁷ Coombes et al., ²⁸ Smidt et al., ²⁹ and various regimen protocols were followed from existing literature in the study. Bissett et al. ²⁷ and Smidt et al. ²⁹ followed the same exercise protocol as per previous literature. Coombes et al. ²⁸ followed the protocol advised by the previous study of Coombes et al. Tonks et al. referred to another study ³⁷ for exercise protocol. Other studies didn’t focus on a specific exercise regimen.

Corticosteroid injection

Drug type

The use of corticosteroid injection was also variable concerning drug, dosage, frequency, and site of the injection. In study^{21,24,27–30} used 10 mg/mL of triamcinolone derivative. Studies 5 ³¹ and 11 ²² used 40 and 20 mg/mL of triamcinolone derivatives, respectively. Studies 4 ²⁵ and 6 ²⁰ used 20 mg/mL of prednisolone, whereas Studies 8 ²⁶ and 9 ²³ used 40 mg/mL of prednisolone.

Number of injections and technique

In several studies,^{20,22–25,28,30,31} a single injection application at the most tender site was performed. Variably, the two to three injection application at different follow-ups and technique of application at different tender sites till the pain-free elbow range of motion was achieved in three studies.^20,21,29 In only one study, a single injection application was done with the peppering technique. ²⁶ Therefore, most studies used single injection applications which obviously saved time and resources for healthcare and patient well-being. A meta-analysis compared corticosteroid injection with placebo reporting no difference in pain intensity when applied on the proximal insertion of Extensor Carpi Radialis Brevis Origin at a 6-month follow-up. ³⁸ No study mentioned the use of ultrasound guidance for injection application.

Risk of bias

Cochrane risk of bias assessment revealed the studies to be collective with unknown to low risk of bias.

Study follow-up duration and outcome

In three studies,^27,29,31 CSI performed better than PTI in the sixth week but 52 weeks follow-up reported PTI to be better than both CSI. The study by Coombes et al. ²⁸ reported CSI, PTI, placebo injection wait, and see policy equal at a follow-up of 52 weeks. This surely reminds us of the statement by Cyriax which meant lateral elbow tendinopathy is self-limiting within 1-year follow-up. Lateral elbow tendinopathy resolves within 1 year as told by Cyriax earlier. ⁶

Erpala et al. ²⁴ reported outcomes for a follow-up of 4 weeks where CSI, RMG, and KT were the same.

Koçak et al. ²⁵ told KT + CSI were better than KT and CSI alone, whereas KT alone performed better than CSI in the 12th-week follow-up. Uyghur reported both CSI and DN to have a significant clinical effect in a 6-months follow-up. PRTEE score was slightly better in the case of DN. Uzunca et al. ²³ reported CSI to be better than PTI and Sham PTI in the third-week follow-up to VAS during activity and during Dorsiflexion. In the third month, PTI was better than CSI and Sham PTI with VAS during rest, activity, and night. Although, not much clinical difference was reported within groups.

Verhaar et al. ²¹ reported all outcomes favoring CSI in comparison to PTI in the sixth week. At 1-year follow-up, no significant difference was recorded. Crowther reported CSI to be better than PTI (ESWT) at a 3-month follow-up. Gündüz et al. ²⁰ conducted a study in comparison to CSI and ESWT and reported no significant difference in first, third, and sixth-month follow-ups. Tonks et al. ³⁰ reported CSI to be better than PTI. Also, CSI + PTI was no better than PTI alone. Regarding the recurrence rate of lateral elbow tendinopathy, only three studies reported the recurrence rate.^1,2,10 In the study of Bisset et al., ²⁷ the rate was 72% for the corticosteroid, wait-and-see 9%, and PTI was 8%. Coombes et al. ²⁸ rate CSI alone at 55%, 54% for CSI + PTI, 20% for placebo alone, and 5% for placebo + PTI. Verhaar et al. ²¹ study reported a rate of 14% for PTI and 40% for CSI. Therefore, this suggests PTI has a relatively less recurrence rate than CSI.

Among the included studies symptom duration varied from 2 to 33 weeks with the mean duration being 12 weeks. In a recent meta-analysis by Ikonen et al., ³⁹ the authors mentioned that there is no correlation between the duration of symptoms before trial enrolment with that of treatment response.

Outcomes

The outcome selectivity was quite variable when comparing these 12 studies. Most studies focused on pain in different instances, grip strength, tendon thickness, treatment satisfaction, and disability. Although most outcomes were symptom-centric and patient-rated, many of them were pathology oriented. There is a significant absence of outcomes related to the quality of life (QoL). Only one study reported the outcome values for QoL through Euro-QoL 5 ED (EQ5D) range. ²⁸ Therefore, the need for QoL outcomes is important in any intervention/disease-driven study.

Additional treatment requirement

When talking about a particular treatment approach we usually focus on whether the treatment was optimal for the patient's satisfaction and outcome. Studies have reported more use of additional treatment in the CSI group than in the PTI group. Also, more sessions of PTI were required as part of additional treatment in the PTI group.

Such data reveals CSI to be less efficient when comparing the PTI group. Also, the majority of the PTI patients did require additional treatment sessions which highlight the decision of PTI prescription with considering factors like patient's choice, resource availability, and sessions required. Although most additional treatment types were oral NSAIDs in both groups, there is insufficient evidence to advise or discard the use of oral NSAIDs. ⁴⁰

Complication rate and recurrence rate

There was a relatively high rate of complications in the CSI group in comparison to the PTI group. Although, most of the complications were not much significant and didn’t interfere with treatment delivery and patient satisfaction. Also, there was a higher recurrence rate in the CSI group than in the PTI group which is an important factor when deciding treatment options.

Comments

We highly suggest future researchers structure the study protocol by suggesting the optimal follow-up period, outcome assessment, and intervention characteristics like dose, frequency, and duration. This would provide uniformity in prospective qualitative and quantitative analysis for LE. Due to the recent emphasis on QoL as an outcome for any disease, ⁴¹ it should be given priority in the outcome assessment in future literature for LET. Although, EQ-5D as a quality of life outcome is still questionable for LET. ⁴² Moreover, recent literature reported core set outcomes (COS) for lateral elbow tendinopathy which should be considered in future research. ⁴² Our study analyzed various PTIs in comparison with corticosteroid injection which provided heterogeneity in the results and understanding of which intervention was better. But when considering factors like the conclusion, follow-up results, additional treatment requirement, and recurrence rate, the PTI group shows superiority over the CSI group.

Limitations of study

As per the registered protocol, we were supposed to conduct a meta-analysis using VAS as the primary outcome. However, due to heterogeneity in the conditions of measuring the score, duration of symptoms as well as the follow-up, we were unable to pool the data. Further, there was wide variability in the outcome types, which made it more difficult to consider other outcomes.

Conclusion

Our review suggests corticosteroid injection provides beneficial short-term effects and PTIs provide intermediate to long-term effects in patients with lateral elbow tendinopathy. This conclusion still requires high-quality comparative randomized trials for a better quality of evidence.

Abbreviations

CSI

corticosteroid injection

PTI

physical therapy intervention

KT

kinesiotaping

DN

dry needling

UST

ultrasound therapy

DTFM

deep transverse friction massage

ESWT

extracorporeal shockwave therapy

MET

muscle energy technique

LLLT

low-level laser therapy

LTE

lateral elbow tendinopathy