Online Information About the Effectiveness of Shoulder Surgery Is Not Based on the Best Available Evidence: A Content Analysis

Education can reassure people with rotator cuff-related shoulder pain: a 3-arm, randomised, online experiment

ABSTRACT

We aimed to investigate the immediate effect of best practice education (with and without pain science messages) and structure-focused education on reassurance among people with rotator cuff-related shoulder pain. We conducted a 3-arm, parallel-group, randomised experiment. People with rotator cuff-related shoulder pain were randomised (1:1:1) to (1) best practice education (highlights that most shoulder pain is not serious or a good indicator of tissue damage and recommends simple self-management strategies); (2) best practice education plus pain science messages (which attempt to improve understanding of pain); and (3) structure-focused education (highlighting that structural changes are responsible for pain and should be targeted with treatment). Coprimary outcomes were self-reported reassurance that no serious condition is causing their pain and continuing with daily activities is safe. Secondary outcomes measured management intentions, credibility and relevance of the education, and similarity to previous education. Two thousand two hundred thirty-seven participants were randomised and provided primary outcome data. Best practice education increased reassurance that no serious condition is causing their pain (estimated mean effect 0.5 on a 0-10 scale, 95% confidence interval [CI] 0.2-0.7) and continuing with daily activities is safe (0.6, 95% CI 0.3-0.8) compared with structure-focused education . Adding pain science messages to best practice education slightly increased both measures of reassurance (0.2, 95% CI 0.0-0.4). Clinicians treating patients with rotator cuff-related shoulder pain should highlight that most shoulder pain is not serious or a good indicator of tissue damage and recommend simple self-management strategies. The benefit of adding pain science messages is small.

Diagnostic labels and advice for rotator cuff disease influence perceived need for shoulder surgery: an online randomised experiment

QUESTION

What are the effects of diagnostic labels and advice, and interactions between labels and advice, on perceived need for shoulder surgery for rotator cuff disease?

DESIGN

2×2 factorial online randomised experiment.

PARTICIPANTS

People with shoulder pain.

INTERVENTION

Participants read a scenario describing a patient with rotator cuff disease and were randomised to bursitis label plus guideline-based advice, bursitis label plus treatment recommendation, rotator cuff tear label plus guideline-based advice, and rotator cuff tear label plus treatment recommendation. Guideline-based advice included encouragement to stay active and positive prognostic information. Treatment recommendation stressed that treatment is needed for recovery.

OUTCOME MEASURES

Perceived need for surgery (primary outcome), imaging, an injection, a second opinion and to see a specialist; and perceived seriousness of the condition, recovery expectations, impact on work performance and need to avoid work.

RESULTS

A total of 2,024 responses (99.8% of 2,028 randomised) were analysed. Labelling as bursitis (versus rotator cuff tear) decreased perceived need for surgery (mean effect -0.5 on a 0-to-10 scale, 98.3% CI -0.7 to -0.2), imaging and to see a specialist, and perceived seriousness of the condition and need to avoid work. Guideline-based advice (versus treatment recommendation) decreased perceived need for surgery (mean effect -1.0, 98.3% CI -1.3 to -0.7), imaging, an injection, a second opinion and to see a specialist, and perceived seriousness of the condition and recovery expectations. There was little to no evidence of an advice label interaction for any outcome.

CONCLUSION

Labels and advice influenced perceived need for surgery and other secondary outcomes in people with rotator cuff disease, with larger effects for advice. There was evidence of little or no interaction between labels and advice for any outcome, but the additive effect of labels and advice appeared large for some outcomes (eg, perceived need for imaging and perceived seriousness of the condition).

TRIAL REGISTRATION

ACTRN12621001370897.

Overview of the Drivers of Low-Value Care Comment on "Key Factors that Promote Low-Value Care: Views of Experts From the United States, Canada, and the Netherlands"

Verkerk and colleagues explored the key drivers of low-value care from the perspective of 18 policy-makers and researchers who had led and evaluated at least one initiative to reduce low-value care or had been responsible for reducing low-value care in an organisation. They identified several drivers of low-value care presented in the 2017 Lancet Right Care Series (eg, fee for service payment systems, the pharmaceutical and medical device industry, fear of malpractice litigation, issues with research conduct and reporting, a culture of 'more is better' and 'new technology is better') but did not discuss some other important ones. In this commentary, we aim to extend the work of Verkerk and colleagues and provide some additional perspectives on the drivers of low-value care within the following categories: Economic incentives; Money, finance, and organisation; Knowledge beliefs, assumptions, bias and uncertainty; and Power and human relationships.

Who writes this stuff? Musculoskeletal information quality and authorship of popular health websites: A systematic review

BACKGROUND

Highly trafficked health websites are major sources of information, but the quality of their musculoskeletal information has not been thoroughly evaluated or their authorship characterized.

OBJECTIVES

To review information about common musculoskeletal conditions on highly trafficked websites and characterize their credibility, authorship, accuracy of information (as compared to treatment guidelines), and consistency with best practice recommendations.

DESIGN

Systematic review.

METHODS

We reviewed the top 15 most highly trafficked health websites, identified by web traffic data. Information about 7 common musculoskeletal conditions was identified and data extracted. Credibility was assessed using the Trust It or Trash It? tool, author backgrounds were identified, accuracy was determined by comparing webpage treatment recommendations to guidelines or systematic reviews, and consistency with best practice recommendations was assessed.

RESULTS

Of 1760 webpages screened, 87 were reviewed. Less than half (44.8%, 39/87) had appropriate sources listed, but 65.5% (57/87) were updated in the previous 5 years. Journalists authored most webpages (55.2%, 48/87). Physician involvement was mostly editorial, and they often lacked expertise in musculoskeletal conditions. Information accuracy was concordant with guidelines for 49.4% (43/87) of webpages, but varied by condition. About half of best practice recommendations were followed (49.1%, 427/870). Pages were unlikely to mention psychosocial factors (16.1%, 14/87), limitations of imaging (18.4%, 16/87), or staying at work (4.6%, 4/87).

CONCLUSIONS

Popular health websites scored poorly for credibility, accuracy, and consistency with best practice recommendations for musculoskeletal conditions. Authorship, bias, and unsupported information are potential sources of inaccuracies that should be addressed in future by these websites.

Online information about the management of anterior cruciate ligament ruptures in Australia: A content analysis

BACKGROUND

Most people who suffer an anterior cruciate ligament (ACL) injury search for information online.

OBJECTIVES

Summarise the proportion of webpages on ACL rupture management that present evidence-based information.

DESIGN

Content analysis.

METHODS

We examined webpage information on ACL ruptures identified through (1) Google searches using terms synonymous with 'anterior cruciate ligament rupture' and searching 'knee surgeon' linked to each Australian capital city, and (2) websites of professional associations. The primary outcome was the proportion of webpages that suggest people can return to at least some form of sport with non-surgical management. Secondary outcomes included webpage information on return to sport with ACL reconstruction (ACLR) and non-surgical management, benefits, harms, and risk of osteoarthritis related to these options, and activity modification.

RESULTS

Out of 115 webpages analysed, 48% suggested people can return to at least some form of sport with non-surgical management. Almost half of webpages suggested most people will return to some form of sport following ACLR (41%) and mentioned benefits of ACLR (43%). Fewer webpages mentioned benefits of non-surgical management (14%), approximately two in three people return to pre-injury level of sport following ACLR (4%), risk of re-injury following ACLR (23%), most people return to sport within 9 months of ACLR (27%), activity modification as a management approach (20%), and ACLR will reduce the risk of osteoarthritis (23%).

CONCLUSION

Most online information on ACL rupture management isn't aligned with the best available evidence. Inaccurate information could mislead patients' treatment choices and create unrealistic expectations for return to sport.