Current evidence for nonpharmacological interventions and criteria for surgical management of persistent acromioclavicular joint osteoarthritis: A systematic review

Abstract

Background

The primary aim of this systematic review was to investigate the individual/combined effectiveness of nonpharmacological interventions in individuals with persistent acromioclavicular joint osteoarthritis. The secondary aims were to investigate the comparative effectiveness of nonpharmacological versus surgical interventions, and to identify the criteria used for defining failure of conservative interventions in individuals who require surgery for persistent acromioclavicular joint osteoarthritis.

Method

Major electronic databases were searched from inception until October 2018. Studies involving adults aged 16 years and older, diagnosed clinically and radiologically with isolated acromioclavicular joint osteoarthritis for at least three months or more were included. Studies must explicitly state the type and duration of conservative interventions. Methodological risk of bias was assessed using the Modified Downs and Black checklist.

Results

Ten surgical intervention studies were included for final synthesis. No studies investigated the effectiveness of nonpharmacological interventions or compared them with surgical interventions. Common nonpharmacological interventions trialed from the 10 included studies were activity modification (n = 8) and physiotherapy (n = 4). Four to six months was the most often reported timeframe defining failure of conservative management (range 3–12 months).

Conclusions

Currently, there is no evidence to guide clinicians about the individual or combined effectiveness of nonpharmacological interventions for individuals with persistent acromioclavicular joint osteoarthritis.

Introduction

Persistent shoulder pain is a major disabling condition affecting up to 30% of the general population.¹ Pathology of the acromioclavicular joint (ACJ) accounts for 12.7–24% of all patients with shoulder pain.^2,3 Osteoarthritis of the ACJ is reported to be one of the most common pathologies of the shoulder joint; however, the exact incidence of isolated ACJ osteoarthritis in symptomatic populations is unknown.^4,5 It is estimated to be 7%.⁵ The first presentation of ACJ osteoarthritis is generally in the fourth decade of life, although it may occur as early as the second decade.^6,7

Osteoarthritis of the ACJ can lead to pain and functional impairment when performing manual labor tasks, sporting activities, and activities of daily living.^5,8–10 This is of particular concern when the symptomatic joint is on the individual's dominant limb.⁵ Physiotherapy is often the first line of conservative management for shoulder complaints, with 53–79% of general practitioners referring their patients with persistent shoulder pain to physiotherapists.^11,12 Conservative management for people with persistent ACJ osteoarthritis consists of either nonpharmacological or pharmacological interventions. Nonpharmacological interventions include, but are not limited to, physiotherapy modalities (e.g. manual therapy, electrotherapy, and exercise therapy), ice, heat, and activity modification.^4,5,13–15 Pharmacological interventions are either injection therapy and/or drug therapy.^4,5,13–15

Current clinical practice suggests that conservative management will yield positive results and should be employed as a first line of management in individuals with ACJ osteoarthritis.^{4,5,8,13–26} Despite this, the evidence for conservatively managing these clients with nonpharmacological interventions is scarce and primarily based on clinical opinion. To date, only a case report has been identified that has investigated the effects of a nonpharmacological intervention (i.e. manual therapy) for the management of isolated ACJ osteoarthritis.²³ Although the case report reported improved shoulder internal rotation and reduced pain at rest and with activity following six sessions of passive mobilization and stretching of the shoulder joint, the results need to be interpreted with caution because of its study design and subsequent low level of evidence.

Surgical intervention is the treatment of choice on failure of conservative management.^{5,15,17–19,21,27,28} Common surgical interventions for people with persistent ACJ osteoarthritis include arthroscopic distal clavicle resection (direct superior approach or indirect subacromial approach) or open distal clavicle resection.^13,14,27 However, the criterion defining the failure of conservative management and the best practice evidence for existing nonpharmacological interventions is currently unknown. For clinicians to implement evidence-informed practice in the management of individuals with persistent ACJ osteoarthritis, it is important to understand the most effective nonpharmacological interventions and the criteria for defining failure of conservative management and the point of referral to a surgeon. Therefore, the main aims of the review were as follows:

1. To investigate the individual/combined effectiveness of nonpharmacological interventions for the management of individuals with persistent ACJ osteoarthritis.

2. To investigate the comparative effectiveness of nonpharmacological versus surgical interventions in individuals with persistent ACJ osteoarthritis.

3. To identify the criteria used for defining failure of conservative interventions in individuals who require surgery for persistent ACJ osteoarthritis.

Methods

Protocol and registration

The review protocol was registered a priori with the International prospective register of systematic reviews, PROSPERO (CRD42018105677).²⁹ The review was reported in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.³⁰

Eligibility criteria

Studies were included based on the following criteria:

Participants

1. Males or females aged 16 years and older diagnosed with isolated persistent ACJ osteoarthritis using findings from a clinical and radiological assessment.^15–17 Common features from the subjective assessment include pain localized to the ACJ that may extend into the trapezius or anterior deltoid and is aggravated by activities such as lifting weights or reaching across your body.^31,32 Common features of the objective examination include localized swelling and tenderness at the ACJ, a positive cross body abduction stress test, acromioclavicular resisted extension test, active compression test (O'Briens test), or Paxinos test.^31,32 The clinical diagnosis can be supplemented by radiographs that commonly show hypertrophic arthritic changes within the ACJ and can rule out other causes of shoulder pathology.^4,27,33–35

2. Symptoms present for at least three months prior to the intervention.³⁶

3. In studies including participants with other shoulder conditions, the proportion of participants diagnosed with isolated ACJ osteoarthritis must be greater than 10%. Due to the low incidence of isolated ACJ osteoarthritis,⁵ studies inclusive of alternative diagnoses were also included in the review to incorporate a broader scope of the evidence.

Interventions

Participants treated using conservative interventions for a minimum of three months prior to surgery were included. For the purpose of this review, we have defined conservative intervention as including both nonpharmacological and pharmacological interventions.

1. Nonpharmacological: Manual therapy, electrotherapy, exercise therapy, ice, heat, education, self-management strategies, and activity modification

2. Pharmacological: Injection therapy and drug therapy

3. Surgical: Arthroscopic distal clavicle resection and open distal clavicle resection

Comparators

Studies comparing nonpharmacological intervention(s) in participants with persistent ACJ osteoarthritis to any of the following intervention(s) have been included: surgical, pharmacological, nonpharmacological, placebo, and no intervention.

Outcomes

Standardized outcome measures measuring pain-related disability, functional outcomes, and pain severity were included. Common examples include the Shoulder Pain and Disability Index,^37,38 American Shoulder and Elbow Surgeons Score,^37,39 Global Rating of Change Scale,⁴⁰ and the Visual Analogue Scale.⁴¹

Study design

All study designs were included.

Studies were excluded based on the following criteria:

1. Not differentiating participants with isolated ACJ osteoarthritis from other ACJ pathologies (osteolysis, os acromiale, dislocation, sprain, fracture, instability, Ewing's sarcoma in children, lymphoma or myeloma in adults, rheumatoid arthritis, septic arthritis, gout, pseudo gout, cysts, ganglia) or other causes of shoulder pain (referred pain from visceral sources, referred pain from spinal sources, subacromial impingement syndrome, adhesive capsulitis, calcific tendinopathy, glenohumeral osteoarthritis, rotator cuff tears)^9,10,27

2. Nonhuman studies (e.g. animal studies and cadaveric studies).

3. Non-English and nonpeer reviewed studies (e.g. conference proceedings).

Information sources

The electronic databases Medline (1946–present via OVID), AMED (1985–present via OVID), PsycINFO (1806–present via OVID), Cochrane Library (1991–present via Wiley), PubMed (1960–present), CINAHL (1981–present via EBSCOhost), Scopus (1823–present), and Google Scholar (2004–present) were searched by the lead investigator (GF) from their inception until 7 August 2018. All references were exported to data management software (Endnote X7—Thomson Reuters; Philadelphia, Pennsylvania). The following search was updated before the final analyses on 7 October 2018 and any further eligible studies were retrieved for inclusion.

Search

Medical Subject Heading terms, specific key words and phrases were searched in MEDLINE, AMED, PsycINFO, PubMED, and CINAHL. Only key words and phrases were searched for in Scopus, Google Scholar, and Cochrane Library. Searches were conducted using Boolean operators, truncation, phrase searching, and synonyms. A subject librarian peer-reviewed the search strategy on multiple occasions. An in-text example of the search strategy for MEDLINE (via OVID) is shown in Table 1. The search strategy for each individual database has been tabulated in Supplementary Appendix 1.

Table 1.

MEDLINE (via OVID) search strategy.

Subject heading databases	Subject headings and key words used	Results
Medline (via OVID)	01. exp Acromioclavicular Joint/ 02. exp Shoulder Joint/ 03. exp SHOULDER/ or exp SHOULDER PAIN/ or exp SHOULDER INJURIES/ 04. exp CLAVICLE/ 05. exp ARTHRITIS/ 06. exp Musculoskeletal Manipulations/ 07. exp CONSERVATIVE TREATMENT/ 08. exp CRYOTHERAPY/ 09. exp Pain Management/ 10. exp Physical Therapy Modalities/ 11. exp Physical Therapy Specialty/ 12. exp Sports Medicine/ 13. exp ORTHOPEDICS/ 14. exp ARTHROSCOPY/ 15. exp mumford.mp. 16. exp distal clavicle excision.mp. 17. exp. Distal clavicle resection. mp. 18. 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 or 17 19. 1 or 2 or 3 or 4 20. 5 and 18 and 19 21. limit 20 to (English language and humans)	299

Study selection

After duplicates removal, GF screened titles and abstracts against the predetermined eligibility criteria. Two reviewers (GF and HD) independently screened full texts and any discrepancies between the reviewers were settled by mutual discussion. A third investigator (LW) was available for arbitration if necessary. GF conducted a secondary search from both the references and citations of the included articles.

Data extraction

The reviewers mutually agreed upon the data to be extracted and formulated an a priori data extraction form. The following data were extracted from the included studies: study (country), design, participants (number, gender, age, duration of ACJ pain, percentage with isolated ACJ osteoarthritis), intervention(s), outcomes(s), conservative intervention(s) trialed, and minimum timeframe for trialing conservative intervention(s).

GF extracted the data from the included studies, which were later verified by HD. Any discrepancies were discussed until a consensus was reached.

Assessment of risk of bias

Due to the heterogeneity in study design of the included studies, the Modified Downs and Black checklist was used to assess the methodological quality (Supplementary Appendix 2).⁴² The modified version has a high inter-rater (r = 0.75) and test–retest (r = 0.88) reliability.⁴² The risk of bias assessment was conducted by GF and later verified by HD. Inconsistencies regarding risk of bias assessment were settled by discussion between the two reviewers (GF, HD). A third author (LW) was available to resolve any disagreements.

Synthesis of results

A meta-analysis of the included studies results was not possible due to heterogeneity in study design and outcome measures used from the included studies. Subsequently, a narrative synthesis was used to summarize the results.

Results

Study selection

The electronic search of the databases yielded 1535 articles (Figure 1). Two additional studies that met the criteria for inclusion were identified from checking the references and citations of the included studies after full text screening.^21,22 Thirty-three potentially relevant studies were identified for full text evaluation. After full text screening, 10 articles were included for final review.^{15–17,20–22,43–46} An updated search found no further eligible studies for inclusion.

Figure 1.

PRISMA flow diagram of the selection process to identify relevant studies.

Study characteristics

The demographic and study characteristics are summarized in Table 2. Of the 10 included studies, two were randomized controlled trials,^15,21 one was a retrospective cohort study,²² one was a prospective case series,⁴⁶ and six were retrospective case series.^{16,17,20,43,45,47} The number of participants in each study ranged from 10 to 103, and the mean age of participants was 42 years (range 28–52.5). The average duration of ACJ pain until initial conservative treatment was reported in two studies,^16,17 whereas the duration of ACJ pain was implied from the minimum timeframe required to trial conservative management from the remaining eight studies.^{15,20–22,43,45–47} The percentage of participants with isolated ACJ osteoarthritis ranged from 32 to 100%. Participants were recruited from diverse populations including: athletic populations,¹⁵ military populations,²¹ and mixed populations (nonathletes, recreational athletes, overhead manual workers, desk job workers and manual laborers).^17,20,46 Five studies did not report the source of population.^{16,22,43–45}

Table 2.

Demographic characteristics, study characteristics, and results of the included studies (n = 10).

Study (country)	Design (level of evidence)	Participants -N (gender, shouldera) -Age mean (range) -Duration of ACJ pain (months)b -Diagnosis N (%)c	Intervention	Outcome	Conservative intervention(s) trialed	Minimum timeframe for trialing conservative intervention(s) N (months)
Charron et al.15 (United States)	Randomized controlled clinical trial (II)	-34 (29 M, 5 F) -28 (16–44) -No data -12 (32.0)	Arthroscopic resection of the distal clavicle (direct superior approach versus indirect subacromial approach)	-ASES -Athletic Shoulder Scoring System	-Nonsteroidal anti-inflammatory drugs -Steroid injections into the joint (ACJ) -Activity modification -Shoulder rehabilitation program consisting of range of motion and rotator cuff strengthening	4–6
Elhassan et al.16 (United States)	Retrospective case series (IV)	-103 (47 M, 56 F, 105) -45 (23–73) -14.4 (3–50)d -48 (46.0)	ACJ resection (arthroscopic versus open resection)	-Postoperative patient satisfaction (Y/N) -Pain level over the ACJ with cross body abduction testing (VAS postoperative) -Tenderness to palpation of the ACJ (Y/N) -Subjective shoulder value -Constant score and relative (age and gender matched) constant score	-Nonsteroidal anti-inflammatory medications -Physical therapy -Activity modification	3
Flatow et al.20 (United States)	Retrospective case series (IV)	-41 (31 M, 10 F) -32 (17–69) -No data -16 (39.0)	Arthroscopic resection of the distal clavicle (direct superior approach)	-Postoperative pain -Range of motion -Function -Rated as either excellent, good, or poor	-Nonsteroidal anti-inflammatory medication -Heat -Activity modification -ACJ steroid injections	10.5
Freedman et al.21 (United States)	Randomized prospective clinical trial (II)	-17 (16 M, 1 F) -40 (24–56) -No data -8 (47.1)	Distal clavicle excision (arthroscopic versus open excision)	-VAS (primary outcome measure) -Bodily pain and physical functioning subscale scores on the SF36 (secondary outcome measure) -Modified ASES (secondary outcome measure) -Satisfaction questions about work and activity status (secondary outcome measure)e	-Cryotherapy -Nonsteroidal anti-inflammatory drugs -Activity modification -Other treatment modalities (not specified)	6
Gartsman17 (United States)	Retrospective case series (IV)	-20 (13 M, 7 F) -51.3 (33–61) -14.3 (6–48) -20 (100)	Arthroscopic resection of the ACJ	-ASES for pain, activities of daily living, work, and sports	-Nonsteroidal anti-inflammatory medication -Avoidance of painful positions -Interdiction of activities that elicited pain -Passive range of motion exercises at home -Local application of heat -Gentle stretching and warm-up exercises before risky activities -Application of ice if there was pain after exertion -Changes at work and in home to minimize shoulder level and overhead activity -Selective strengthening of the rotator cuff and three parts of the deltoid using surgical tubing to provide progressive resistance and performed with the arm less than 90 ° of abduction -Trapezius and serratus anterior strengthening -Intra-articular steroid injections -Oral cortisone	12
Gokkus et al.43 (Turkey)	Retrospective case series (IV)	-30 (12 M, 18 F) -52.5 (22–69) -No data -30 (100)	Open limited distal clavicle excision	-VAS -UCLA	-Conservative treatment consisted of three steps 1. Activity modification, paracetamol, and/or nonsteroidal anti-inflammatory drugs for three months 2. 15-day regimen requiring regular hospital visits consisting of superficial and deep tissue heating with ultrasonography, electrotherapy (low frequency currents), manual therapy, and shoulder muscle strengthening exercises 3. Steroid (methylprednisolone) with lidocaine injection	3.5f
Kay et al.46 (United States)	Prospective case series (IV)	-10 (8 M, 2 F) -44 (29–52) -No data -5 or 6 (50.0–60.0)g	Arthroscopic distal clavicle excision	-UCLA	-Nonsteroidal anti-inflammatories -Narcotics	4
Levine et al.44 (United States)	Retrospective case series (IV)	-60 (45 M, 15 F, 66) -46 (21–78) -No data -58 (88.0)h	Arthroscopic distal clavicle resection (direct superior approach versus indirect subacromial approach)	-ASES -Subjective pain (VAS) -Subjective instability (VAS) -Patient satisfaction (Y/N) -Patient recommendation (Y/N) -Failure (Y/N)	-Nonsteroidal anti-inflammatory drugs -Physical therapy -Activity modification	6
Novak et al.45 (United States)	Retrospective case series (IV)	-21 (23 shoulders)i -36 (21–49) -No data -10 (43.5)	Open distal clavicle resection	-Modification of the HSS (involvement of litigation or workers compensation was noted) -Isokinetic muscle strengthj	-Nonsteroidal anti-inflammatory agents -Ice -Activity modification -Physical therapy	6
Robertson et al.22 (United States)	Retrospective cohort study (III)	-48 (25 M, 23 F, 49) -49 (19–76) -No data -44 (91.7)	Distal clavicle excision (arthroscopic versus open resection)	-ASES -VAS -Surgical time -Minimum radiographic acromioclavicular joint distance -Subjective shoulder value (percentage of normal) -Scar satisfaction (Y/N) -Willingness to have the surgery again (Y/N)	-Nonsteroidal anti-inflammatory drugs -Steroid injection -Physical therapy	6

ACJ: acromioclavicular joint; ASES: American Shoulder and Elbow Surgeons score; F: female; HSS: Hospital for Special Surgery (HSS) shoulder assessment; M: male; SF36: Short Form 36; UCLA: University of California Los Angeles Shoulder rating scale; VAS: Visual Analog Scale; Y/N: Yes/No.

^aTotal number of shoulders operated on if different to the total number of participants.

^bDuration of ACJ pain to initial conservative treatment mean (range).

^cDiagnosis of isolated ACJ osteoarthritis. If the included study did not state the percentage of participants with isolated ACJ osteoarthritis, it was manually calculated from the studies' raw data.

^dUnsure if data are duration of ACJ pain to initial conservative treatment or surgical intervention.

^eNo data on how this outcome was measured.

^fConservative intervention trialed for at least three months and 15 days, with a trial of pharmacological intervention after this time period for an unspecified period of time.

^gIsolated ACJ osteoarthritis is present in 50–60% of participants because it was not specified whether four participants with impingement had osteoarthritis or osteolysis.

^hFifty-eight (88%) of the shoulders operated on had radiographic ACJ osteoarthritis. However, 37 (56%) of the shoulders also had impingement. This 37 (56%) was from the total population, not the radiographic ACJ osteoarthritis population, so the percentage of isolated ACJ osteoarthritis could not be determined with full accuracy

ⁱNo data were given on the proportion of males or females.

^jOnly assessed in participants with unilateral shoulder involvement n = 18/21.

Interventions

No studies investigated the outcomes of nonpharmacological interventions and all the included studies investigated the outcomes of surgical interventions.^{15–17,20–22,43–46} The type of surgical interventions investigated has been presented in Table 2. One surgeon performed most of the surgeries.^{15–17,22,43–45} In some studies, only the surgery location(s), setting(s), or date range(s) were stated.^20,46 However, in one study none of the above were reported.²¹

Outcomes

The common outcomes included American Shoulder and Elbow Surgeons score,^{15,17,21,22,44} Visual Analogue Scale,^{16,21,22,43,44} and the University of California Los Angeles Shoulder rating scale^43,46 (Table 2).

Risk of bias within studies

The methodological quality of the included studies is presented in Table 3. The total quality scores ranged from 63.0 to 78.6%. Studies scored well in the reporting domain, whereas studies scored poorly in the external validity domain. Most of the included studies scored poorly for the power calculation question.

Table 3.

Methodological quality of the included studies (n = 10).^a

Modified Downs and Black checklist
	Reporting											External validity				Internal validity—Bias								Internal validity—confounding (selection bias)							Power	Total %
Study	1	2	3	4	5	6	7	8	9	10	%	11	12	13	%	14	15	16	17	18	19	20	%	21	22	23	24	25	26	%	27	%
Charron et al.15	1	1	1	1	2	1	1	1	1	1	100	0	0	0	0	0	0	1	1	1	1	1	71.4	1	1	0	0	0	1	50.0	0	67.9
Elhassan et al.16	1	1	1	1	NA	1	1	1	1	1	100	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	72.2
Flatow et al.20	1	1	1	1	NA	1	1	1	1	1	100	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	72.2
Freedman et al.21	1	1	1	1	2	1	1	1	1	1	100	0	0	0	0	0	0	1	1	1	1	1	71.4	1	1	1	1	0	1	83.3	1	78.6
Gartsman17	1	1	1	1	NA	1	1	1	1	0	88.9	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	66.7
Gokkus et al.43	1	1	1	1	NA	1	1	1	1	1	100	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	72.2
Kay et al.46	1	1	1	1	NA	1	1	1	1	1	100	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	72.2
Levine et al.44	1	1	1	1	NA	1	1	1	1	0	88.9	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	66.7
Novak et al.45	1	1	1	1	NA	1	1	1	1	1	100	0	0	0	0	NA	0	1	NA	1	NA	1	75.0	NA	NA	NA	NA	NA	1	100	0	72.2
Robertson et al.22	1	1	1	1	1	1	1	1	1	1	90.1	0	0	0	0	0	0	1	0	1	NA	1	50.0	1	1	0	0	0	1	50	1	63.0

NA: not applicable.

^aItems 5, 14, 17, 19, 21, 22, 23, 24, 25 were removed from the Modified Downs and Black checklist as most of the included studies were cross-sectional studies.

Synthesis of results

No studies were identified investigating the effectiveness of nonpharmacological interventions and the comparative effectiveness with surgical interventions in individuals with persistent ACJ osteoarthritis. However, the common nonpharmacological interventions trialed include activity modification,^{15–17,20,21,43–45} physiotherapy,^16,22,44,45 exercise therapy,^15,17,43 and ice therapy^17,21,45 (Figure 2).

Figure 2.

Frequency of conservative interventions trialed between the included studies (n = 10).

The frequency of conservative interventions (including pharmacological interventions) trialed between the included studies has been depicted in Figure 2. The most common conservative interventions trialed in 8 out of the 10 studies were activity modification^{15–17,20,21,43–45} and oral nonsteroidal anti-inflammatory drugs (NSAIDs),^{15–17,21,22,44–46} and in 4 out of the 10 studies were intra-articular steroid injections^15,17,20,22 and physiotherapy.^16,22,44,45 The least common conservative interventions trialed were avoidance of painful activities or positions,¹⁷ combined steroid and lidocaine injection,⁴³ electrotherapy,⁴³ manual therapy,⁴³ narcotics,⁴⁶ oral cortisone,¹⁷ paracetamol,⁴³ ultrasonography,⁴³ and warm-up exercises before any risky activities.¹⁷

The minimum timeframe for trialing conservative intervention(s) before surgery could be between 0 and 12 months as shown in Figure 3. Four to six months were the most often reported timeframe, as reported in six studies.^{15,21,22,44–46}

Figure 3.

Minimum timeframe for trialing conservative intervention(s) before surgical intervention (n = 9).^a

^aNo data on the duration of conservative intervention(s) trialed.⁴³

Discussion

Main results

Currently there is no evidence to inform clinicians on the effectiveness of nonpharmacological interventions or the comparative effectiveness of nonpharmacological versus surgical interventions in individuals with persistent ACJ osteoarthritis. Despite this, various studies have used nonpharmacological interventions as a treatment modality.^4,5,13–22 Our review has shown the most commonly used nonpharmacological interventions were activity modification,^{15–17,20,21,43–45} physiotherapy,^16,22,44,45 exercise therapy,^15,17,43 and ice therapy.^17,21,45 Four to six months were the most often reported timeframe for trialing conservative intervention(s) before surgery was indicated.^{15,21,22,44–46}

Although many nonpharmacological interventions were reported to be trialed prior to surgery,^{15–17,20–22,43–45} none of them were tested for effectiveness either in isolation or combination in this clinical population. To the authors knowledge, only one study has specifically investigated the management of ACJ osteoarthritis with nonpharmacological interventions.²³ The primary interventions in this case report were passive mobilization and stretching of the glenohumeral joint, which in turn was hypothesized to alleviate symptoms by reducing compensatory motion of the ACJ by restoring glenohumeral joint internal rotation.²³ As we were unable to determine the duration of ACJ osteoarthritis of the participant, this case report was not included in the current review. However, the findings suggest treating the glenohumeral joint could indirectly influence the symptoms of ACJ osteoarthritis. Despite not specifically investigating the effectiveness of nonpharmacological interventions, one of the included studies from the current review reported that 73% of participants with ACJ osteoarthritis had a positive effect to a trial of conservative management and subsequently did not require surgery.⁴³ Conservative management used in that study consisted of activity modification, drug therapy, ultrasonography, electrotherapy, manual therapy, exercise therapy, and steroid injections.⁴³ Although these two studies provide anecdotal evidence for improved shoulder functioning following conservative management, further research is required to elucidate the mechanisms underpinning the effectiveness of various conservative interventions reported in this review.

The lack of eligible studies for the primary aim of our review is concerning. Possible reasons for this paucity of studies could include that ACJ osteoarthritis may be overlooked²⁷ as it is not common in isolation and often coexists with other more prevalent shoulder pathologies.^21,44,48 It has been suggested that osteoarthritis of the ACJ can occur with pathology of the supraspinatus tendon, and osteophytes from the arthritic joint may contribute to subacromial impingement.^17,49–51 A recent prospective cohort study showed that pathology of the ACJ and sternoclavicular joint was present in 17.7% of participants with subacromial impingement syndrome and rotator cuff disease.⁵² A systematic review⁵³ and a randomized controlled trial⁵⁴ have both shown that exercise therapy has a positive effect on reducing pain-related disability in participants with either rotator cuff pathology or subacromial impingement syndrome. Six of the included studies in our review reported using exercise therapy to treat persistent ACJ osteoarthritis.^{15–17,22,44,45} Therefore, it is plausible that clinicians are using evidence available for other common and coexisting shoulder pathologies to treat persistent ACJ osteoarthritis.

No studies have been identified that compared the effectiveness of nonpharmacological interventions versus surgical interventions in individuals with persistent ACJ osteoarthritis.^8,49 The current evidence for surgical intervention as a treatment option for persistent ACJ osteoarthritis has been derived from studies comparing the efficacy of a single surgical technique or the effectiveness of different surgical techniques.^8,14,55 Recently the role of surgical management for rotator cuff tears and subacromial impingement syndrome has been challenged.^56–58 Shoulder surgery has been challenged in favor of nonpharmacological interventions consisting mainly of exercise therapy due to studies showing that surgical intervention has no difference in outcomes when compared to nonpharmacological interventions and the potential for postsurgical complications.^56–58 Similarities can be drawn between surgical interventions for different shoulder disorders and ACJ osteoarthritis, where poor long-term results of ACJ surgery were reported in some studies and there is the potential for long-term postsurgical complications such as ongoing weakness, instability, and pain.^4,28,33,59 While the effectiveness of different surgical interventions are well researched in individuals with isolated ACJ osteoarthritis,^8,14,55 due to lack of research comparing the clinical effectiveness of surgical versus nonpharmacological interventions, we cannot comment on the comparative effectiveness of surgical and nonpharmacological interventions in individuals with ACJ osteoarthritis.

The lack of studies investigating the effectiveness of nonpharmacological interventions could be due to difficulty in diagnosing isolated ACJ osteoarthritis. A surgical study found that 50% (n = 17) of those diagnosed with isolated ACJ osteoarthritis from a clinical and radiological assessment had concomitant glenohumeral joint pathology arthroscopically that was not picked up in either preoperative assessment.²¹ Reasons for this difficulty in diagnosing isolated ACJ osteoarthritis may be due to the common clinical tests being specific for ACJ pathology (specificity of cross body adduction stress test; acromioclavicular resisted extension test; active compression test is 79, 85, and 95%, respectively) inclusive of alternative diagnoses such as ACJ osteolysis or instability.³¹ Because these tests are used to diagnose ACJ pathology as a collective, the clinical diagnosis of isolated ACJ osteoarthritis remains challenging. Isolated ACJ osteoarthritis may also be difficult to diagnose due to its high prevalence in asymptomatic populations^60,61; MRI imaging studies have shown that ACJ osteoarthritis is present in 76–82% of asymptomatic participants.^60,61

It seems to be common clinical practice to trial both nonpharmacological and pharmacological interventions for at least 4–6 months before conservative management is deemed to have failed and surgical intervention is indicated in individuals with persistent ACJ osteoarthritis.^{15,21,22,44–46} The most common conservative interventions trailed included activity modification,^{15–17,20,21,43–45} oral NSAIDs,^{15–17,21,22,44–46} intra-articular steroid injections,^15,17,20,22 and physiotherapy.^16,22,44,45 Although physiotherapy was reported as an intervention, there was no adequate detail on what it entailed.

Strengths and limitations

Strengths

The review clearly defined the studies to be included, undertook a comprehensive search strategy involving all of the major databases, assessed each studies risk of bias, and was undertaken in accordance with published guidelines by a team of reviewers with more than one member involved in most stages of the review. Every possible attempt was made to locate eligible articles by searching through the reference and citation lists of the included articles.

Limitations

Although this is the first review to investigate the effectiveness of nonpharmacological interventions in people with persistent ACJ osteoarthritis, the results need to be interpreted with the following limitations. First, the included studies' designs were predominately case series with a low level of evidence (IV).^{16,17,20,43,45–47} Due to this, 8 out of the 13 checklist items in the Modified Downs and Black checklist for internal validity were not applicable and there is a high chance the included studies have been subjected to bias. Also from the risk of bias assessment, all studies scored poorly in the external validity section indicating a limited ability to generalize their results. Second, although we included studies with at least 10% of participants with isolated ACJ osteoarthritis, our review included studies with alternative shoulder pathologies, which may have confounded the results. Third, a wider sampling population and criteria for surgery from the included studies meant that there was a limited ability to pool results. Fourth, exclusion of studies published in languages other than English may have subjected this review to publication bias.

Implications for practice

The most commonly used nonpharmacological interventions in clinical practice for persistent ACJ osteoarthritis include activity modification,^{15–17,20,21,43–45} physiotherapy,^16,22,44,45 exercise therapy,^15,17,43 and ice therapy.^17,21,45 However, evidence to support using these interventions is still unknown so clinicians have to be mindful of such limited evidence. It seems that current practice is to trial both nonpharmacological and pharmacological interventions for at least 4–6 months before conservative management is deemed to have failed and surgical intervention is indicated.^{15,21,22,44–46} The most common conservative interventions trailed include activity modification,^{15–17,20,21,43–45} oral NSAIDs,^{15–17,21,22,44–46} intra-articular steroid injections,^15,17,20,22 and physiotherapy.^16,22,44,45 Despite identifying commonly used timeframes and conservative interventions, the evidence to support this common practice is not based on appropriate study designs. Clinicians need to be mindful of the lack of clinical efficacy of such interventions and patients need to be informed in the clinical decision making prior to surgery.

Implications for future research

The lack of evidence investigating nonpharmacological interventions is not the same as evidence suggesting that these interventions are not effective. Future research needs to investigate the effectiveness of nonpharmacological interventions in isolation and combination, and investigate the comparative effectiveness of nonpharmacological versus surgical interventions in individuals with persistent ACJ osteoarthritis. This recommendation is in line with a recent scoping review that critically evaluated the efficacy of conservative and surgical treatments for adults with ACJ pain.⁴⁹ Clearer diagnostic criteria for persistent ACJ osteoarthritis need to be developed so that individuals with isolated clinical presentations can be ascertained and tailored interventions for this clinical population be developed. For the effectiveness of nonpharmacological interventions to be meaningfully compared, it is imperative that future studies use outcome measures validated for ACJ osteoarthritis. Examples include the Specific Acromioclavicular Score or Nottingham Clavicle Score.^62,63

It is also recommended that prescriptive clinical predictive rules be developed from prospective longitudinal studies to determine the most effective combinations of nonpharmacological interventions and the optimal point at which conservative management has failed and surgery is indicated. Further, a Delphi study of clinicians involved in the management of ACJ osteoarthritis could investigate the current practices of diagnosing and managing patients with persistent ACJ osteoarthritis. Results from such studies will create an evidence-informed treatment pathway to guide the management of individuals with persistent ACJ osteoarthritis.

Conclusion

Currently there is no evidence to inform clinicians on the effectiveness of nonpharmacological interventions or the comparative effectiveness of nonpharmacological versus surgical interventions in individuals with persistent ACJ osteoarthritis. Common clinical practice is to trial activity modification, NSAIDs, intra-articular steroid injections, and physiotherapy for at least 4–6 months before conservative management is deemed to have failed and surgical intervention is indicated. This review has identified significant gaps in current practice and highlighted the need for future research to create an evidence-informed treatment pathway for the management of persistent ACJ osteoarthritis.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical Review and Patient Consent

Not applicable.

Guarantor

GF.

Contributorship

GF researched literature and conceived the study. GF and HD were involved in protocol development. GF wrote the first draft of the manuscript. LW critically reviewed and modified the first draft. All authors reviewed and edited the manuscript and approved the final version of the manuscript.

Supplemental material

Supplemental material is available online.