• 
• 

Three Part Question

In [adults with rotator cuff syndrome of either traumatic or non-traumatic cause] is [surgical treatment] when compared to [non-surgical treatment] have a better outcome for patients [return to normal function or an improvement in pain reduction?]

Clinical Scenario

A 38y/o male presents with persistent shoulder pain, after manual clinical assessment and ultrasound it is confirmed he has a rotator cuff tear. The patient expresses concerns and wants to know what his long term treatment options are.

Search Strategy

OVID, PUBMED, CINAHL and COCHRANE used for search input
(((((((((Rotator cuff syndrome[Title]) OR (Rotator cuff tear[Title])) AND (Surgery[Title])) OR (surgical treatment[Title])) AND (Non-surgical[Title])) OR (Physiotherapy[Title])) OR (physical therapy[Title])) OR (conservative management[Title])) OR (full Function[TITLE])) AND (pain[TITLE])

Search Outcome

977 reduced to 27 (pubmed)
4593 OVID
174 reduced to 28 CINAHL

Excluding systematic reviews in order to look at primary research from randomised control trials, 3 results were found appropriate for this PICO with a single Cohort study added to compare.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Ranebo et al 2020 Sweden	58 adult patients without previous shoulder injuries, with reduced flexion and pain following trauma, with full thickness tears not exceeding 2/12 segments of Sagittal MRI of numeral head.	PCRT	Constant-Murley Score (CS)	12 month review CS>70 repair:68.8% Non-operative: 69.2%(P.97) CS>80 repair:59.4% non-operative: 46.2% (P.43) CI 95%	Potential for bias by unblinded physio assessor. 2 patients with unrelated significant injuries may have impacted results (however both have satisfactory CS score>12months) Potential acute-on-chronic tear patients in trial unaccounted for this in results. 3 patients refused MRI and 1 set of patient values missing making at 1 year. Limited power due to patient unavailability.
			Western Ontario Rotator Cuff Index (WORC)	5.0 difference between groups in WORC score, p.62 - showing no significant difference
			Numeric Rating Scale (NRS)	No difference at 12 months between groups for VAS or NRS
Cederqvist et al 03/12/2020 Finland	190 shoulders (187 patients) >35y/o with rotator cuff disease (RCD),after 3months of physiotherapy with persisting subachromial pain	Randomised Trial	Constant-Murley Score (CS)	mean difference 3.4 -0.4 to 7.1 p0.077 CI 95	Uneven treatment between groups - non repair receiving more steroid injections and repair group receiving more supervised pysiotherapy Study used “shoulders” in place of patient classification, leading to 3 patients being repeated. No blinding apparent other than initial randomisation Low exercise attendance/adherence under supervision from non-operative group. Operating surgeon re-assessed interventions in operative group, potentially causing bias
			Visual Analogue Pain Score (VAS)	mean difference 4. VAS -3 to 10 p0.25 CI 95
Moosmayer et al 2014 Norway	103 patients with a rotator cuff tear <3 cm	PRCT	Constant score	5.3 difference at 5 years for surgical group (-0.05 - 10.7)P0.05 CI95%	2 new in juries in non-operative unclear if result of trial or related Surgical group had more physiotherapy than non-operative group: 28 vs 21 sessions. Only took physical component of sf-36 form post initial baseline, which may indicate concerns for mental health impacting recovery. Secondary care recruitment only, may reduce impact on primary care presentation of this group Trial structure hypothesis of surgical crossover may have increased bias without alternative conservative options for treatment either during process or at determined cut off
			American Shoulder and Elbow Score (ASES)	9.0 points greater for surgical group at 5 years -4.2 to 13.8 points (p0.001, CI95%)
			Pain free abduction	14.7 degrees at 5 years in favour of surgical - 0.1-29.4 (p0.07, CI95%)
			Pain free flexion	5.4 degrees in favour of surgical at 5years - 5.4degrees (-4.9 - 15.7)(P0.30 CI95%)
			Strength	0.8kg difference at 5 years (-1.1 - 2.7) (P0.31 CI95%)
Jain et al 13/092019 USA	127 patients >45y/o with symptomatic rotator cuff tears	Multi-centre Cohort Study	Shoulder Pain and DIsability Index (SPADI)	Propensity weighted SPADI difference 22 points. 32.1 surgical 11.8 non-surgical CI95 p<0.01	Non primary research dataset, a systematic review of current research or a randomised control trial would have been more appropriate to examine interventions. Results presented to support hypothesis as >30% or >50% of change in treatment groups Missing MRI for 17 patients Not all data available at each outcome date
			American Shoulder and Elbow Surgeons Standardised form (ASES)	Propensity adjusted ASES difference 22.2 points -32.8 surgical -11.6 non surgical CI95 p<0.01
			Magnetic Resonance Imaging (MRI)

Comment(s)

While there is a case for the evidence presented for RCD management by surgery and conservative means, there is mixed levels of quality between research presented. The evidence displayed furthers the case for physiotherapy and conservative management in the first instance before consideration of surgical intervention. Randomisation was appropriate in all trials however blinding of assessors was an issue and may have increased bias risk. Trial methodology for future primary research should have set times for outcome assessment with enough time for healing to occur. There are difficulties of balancing individual treatment against fair trial methodology. E.g catering physio to individuals to maximise treatment vs a standard set of exercises. Issues like posture, muscle imbalance, age and degenerative musculature and lifestyle may also impact results. Trials have noted and attempted to mitigate this. Clinician influence and approach to patient intervention as well as patient confidence and mental health status should be considered in any future research. Jain et al, (2019) may have had their study influenced by this due to the shared decision making process in inclusion. No trials had placebo surgical interventions, this would provide a control, however ethically is likely inappropriate.

Clinical Bottom Line

Surgical interventions showed greater pain reduction and return to full function than conservative measures. In small rotator cuff tears (or non full thickness tears) the clinical difference was non-significant. This research suggests that conservative management as a first line management in small or non-full thickness tears may be equal to surgical treatment in patients with RCD, with failed conservative management leading to surgery as a second line. Non surgical management may lead to fatty infiltration or degeneration of shoulder, mixed outcome in trials with patients mostly developing coping mechanisms due to pain reduction and others requiring secondary intervention. In large full thickness tears surgery may be more appropriate to reduce pain and increase function. Time to intervention impacts all outcomes regardless of treatment route.

References

1. Ranebo et al, Surgery and physiotherapy were both successful in the treatment of small, acute, traumatic rotator cuff tears: a prospective randomized trial Journal of Elbow and Shoulder Surgey 2020; 459-470
2. Cederqvist et al Non-surgical and surgical treatments for rotator cuff disease: a pragmatic randomised clinical trial with 2-year follow-up after initial rehabilitation Annals of the Rheumatic Diseases 03/12/2020; 796-802
3. Moosmayer et al Tendon Repair Compared with Physiotherapy in the Treatment of Rotator Cuff Tears - A Randomized Controlled Study in 103 Cases with a Five-Year Follow-up Journal of Bone and Joint Surgery 17/09/2014; 1504-1514
4. Jain et al Comparative Effectiveness of Operative versus Non-Operative Treatment for Rotator Cuff Tears: A Propensity Score Analysis from The ROW Cohort American Journal of Sports Medicine 13/09/2019