We welcome comments in the PEDro blog. All comments will be moderated. We only accept comments which are respectful, professional and based on evidence. If a post relates to a research paper, please read the full paper before submitting your comment.
QUADAS, QUADAS-2 and DAQS provide unreliable estimates of quality of studies of diagnostic accuracy in physiotherapy
PEDro is currently working on a new database that will index studies and reviews that evaluate the accuracy of diagnostic tests used by physiotherapists. Called DiTA, this new database project is being led by Mark Kaizik, Rob Herbert and Mark Hancock.
An investigation of the measurement properties of quality assessment tools for diagnostic test accuracy studies was conducted to inform the development of DiTA. The main aims of the investigation were to determine the reliability, measurement error, internal consistency, convergent validity, and floor and ceiling effects of three tools commonly used to evaluate the quality of diagnostic test accuracy studies.
50 diagnostic test accuracy studies in the field of musculoskeletal, orthopaedic or sports physiotherapy that were published in English were randomly selected from DiTA. Three tools were evaluated: Quality Assessment of Diagnostic Accuracy Studies (QUADAS), Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) and Diagnostic Accuracy Quality Scale (DAQS). Two physiotherapists independently rated each study using each tool (the order of both the study and tool were randomised). Summary scores were calculated to facilitate the analyses. 13/14 QUADAS items, 5/7 QUADAS-2 domains, and 14/21 DAQS items had less than moderate inter-rater reliability (Kappa< =0.40). Inter-rater reliability for summary scores ranged from poor (Intraclass Correlation Coefficient 0.27, QUADAS) to moderate (0.45, DAQS). Standard error of measurement was 2.7 points was for the 0-28 point QUADAS tool, 1.8 for the 0-14 point QUADAS-2, and 3.6 for the 0-42 point DAQS. Internal consistency was acceptable (Cronbach’s alpha>0.70) for the QUADAS-2 tool only. Convergent validity was acceptable (Pearson’s correlation>0.70) for half of the analyses: QUADAS vs DAQS (both rater 1 and rater 2), and QUADAS-2 vs DAQS (rater 1). Floor or ceiling effects were not present in any tool. The study concludes that all three tools provide unreliable estimates of quality for studies evaluating the accuracy of diagnostic tests used by physiotherapists.
Kaizik MA et al. Measurement properties of quality assessment tools for studies of diagnostic accuracy. Braz J Phys Ther 2019 Jan 30;Epub ahead of print
Systematic review found that exercise prevents falls in older people living in the community
A recently published Cochrane review evaluates the benefits and harms of exercise interventions for preventing falls in older people living in the community. This review included randomised controlled trials evaluating any form of exercise as a single intervention in people over 60 years old. The primary outcome was rate of falls (falls per person-year) measured at the time point closest to 18 months post-randomisation. Methodological quality of the included trials was evaluated with the Cochrane risk of bias tool, and the quality of the evidence was assessed using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach.
This review included 108 trials with 23,407 participants conducted mostly in high income countries. 77% of included participants were women. The average participant age in the included trials was 76 years. Exercise was compared to a control intervention not thought to reduce the rate of falls in people not recently discharged from hospital in 81 trials (n = 19,684 participants) and in people who were recently discharged from hospital in four trials (n = 816 participants). 53% of the interventions included balance and functional exercises as the primary intervention, followed by three-dimensional training (constant repetitive movement through all three spatial planes; 15% of the interventions).
This review found high-quality evidence from 59 trials (n = 12,981 participants) that exercise interventions reduced the rate of falls by 23% (95% CI 17% to 29%) compared to control intervention not thought to reduce falls. There was low-quality evidence from 10 trials (n = 4,047 participants) that exercise interventions reduced the number of people experiencing a fracture following a fall by 27% (95% CI 5% to 44%) compared to control intervention.
Strong evidence shows that exercise interventions reduce the rate of falls in older people living in the community. Further work is needed to understand the impact of resistance training, dance or walking programs. Larger studies are needed to evaluate the impact of exercise on fall-related fractures and falls requiring medical attention.
Listen to Norman Swan interview Cathie Sherrington (Professor from the Institute for Musculoskeletal Health, University of Sydney who is the lead author of the review) for ABC Radio National’s Health Report.
Sherrington C et al. Exercise for preventing falls in older people living in the community. Cochrane Database Syst Rev 2019 Jan 31;1:CD012424
Read more on PEDro.
#MyPTArticleOfTheMonth – what is France Mourey reading?
France Mourey is a professor in Université de Bourgogne, France who has expertise in geriatric rehabilitation, particularly the assessment of balance and gait, frailty, and training of motor function in Alzheimer’s disease. She coordinates a research program called “Implicit motor learning in Alzheimer’s disease” that is supported by the Agence Nationale de la Recherche (the peak research funding agency in France) and aims to develop virtual reality solutions for people with Alzheimer’s. France is also Vice-Chair of the Espace de Reflexion Éthique Bourgogne - Franche - Comté and chairs the Geriatrics Group of the Société Française de Physiothérapie.
Weber M, et al. Feasibility and effectiveness of intervention programmes integrating functional exercise into daily life of older adults: a systematic review. Gerontology 2018;64:172-187
France chose this article because it evaluates whether exercise that is incorporated into daily routines can improve function and reduce falls in older adults. France says “While structured exercise programs without direct links with daily activities can be very effective in young people, integrating exercises into daily life tasks may be a better approach in older people. This approach addresses the issue of program specificity.”
The systematic review identified six randomised controlled trials comparing integrated training with structured exercise, usual care or inactive control treatment for people aged over 60 years. The trials were conducted in community dwellers with a history of falling, those receiving home-based care, and in institutional care. Meta-analysis could not be performed because of the diversity of the trials. The results of individual trials suggests that integrated training is feasible and may increase adherence and improve some outcomes.
France says “Integrated functional training may be useful for getting older people to exercise. For example, incorporating ankle exercises into daily activities could maintain flexibility, balance and mobility. This review could be used to guide future trials.”
#MyPTArticleOfTheMonth #PhysicalTherapy #physio
#MyPTArticleOfTheMonth resource – how to ask a clinical question
As a clinician, you perform diagnostic tests, provide information on prognosis, and implement interventions on a daily basis. You may want to find out if the diagnostic test (or combination of tests) you are using is the best available considering your facilities and resources. You may want to discover the course of recovery for a condition you don’t see very often. You may also like to know if you are offering intervention that is supported by the results of high-quality research. To do these you need to pose a clinical question.
In order to answer your clinical question, it is helpful to break it down into 4 essential components, using the ‘PICO’ framework. In this memory aid, P stands for patient, problem or population, I stands for intervention, C stands for comparison or control intervention, and O stands for outcome. Taking the time to clearly define the question will help you work out the best search terms to use, which in turn will make finding the best research to answer your question less daunting or time-consuming.
For questions about the effects of interventions, your PICO question should include all 4 elements:
P (patient, problem or population): what is the condition or population group of interest, are you interested in a particular subgroup (eg, acute stroke) or sociodemographic group (eg, workers)? Are you working with older people, children, athletes, people that have had a traumatic brain injury or stroke?
I (the intervention): what treatment are you interested in.
C (the comparison or control intervention): are you interested in comparing your intervention to placebo, usual care, or another intervention (eg, aquatic versus land-based exercise).
O (the outcome): what measurable outcome(s) are you interested in improving? Is the outcome important to patients? Outcomes could be events (eg, falls), symptoms (eg, pain), functional measures (eg, walking speed) and quality of life. Harmful effects and the cost of treatment are also important outcomes to consider.
An example of a PICO question about the effects of intervention is: ‘In older people with knee osteoarthritis, is hydrotherapy more effective than land-based exercise in relieving pain?’
PICO can also be used to frame diagnostic questions, but here “I” takes on a new meaning:
P (patient, problem or population)
I (the “issue”): this could be a diagnostic test, a combination of physical tests, or a clinical prediction rule.
C (the comparison): what do you want to compare your diagnostic test to? This could be a reference test or the gold standard test.
O (for outcome): this is usually a measure of the test utility like specificity or sensitivity. This gives you an idea of both the rate of false positives (diagnosing the condition in those that do not have it) and false negatives (missing the diagnosis in those that do).
An example of PICO question about a diagnostic test is: ‘In female soccer players with knee injuries, what is the accuracy of the anterior draw test compared to medical resonance imaging for detecting an anterior cruciate ligament injury?’
Elements of PICO can help you ask questions about the prognosis of a condition. With prognostic questions “I” takes on a new meaning and the “C” is dropped:
P (patient, problem or population): when specifying this element it is useful to include the duration or severity
I (for “time”): over what time span are you interested in, the short- or long-term?
O (for outcome): these should be both quantifiable and important to patient’s goals and priorities. Examples include the rate of disease progression or a positive outcome (eg, return to work or sport).
An example of a PIO question about prognosis is: ‘For people with an episode of back pain resulting in 4 weeks off work, what is the likelihood that they return to work in their previous role at 6 months?’
PEDro has a great video tutorial on posing clinical questions about interventions. This ‘How to ask a clinical question in PICO format’ video is available in English, French, Portuguese, Spanish, Italian, German, Japanese, Tamil, and Chinese simplified characters.
Support for PEDro comes from industry, physiotherapy organisations and individuals
Support for PEDro comes from industry partners around the globe. The Australian Physiotherapy Association is our Foundation Partner. The Motor Accident Insurance Commission, Transport Accident Commission, Chartered Society of Physiotherapy, and Cerebral Palsy Alliance are Partners. Our Association Partners for 2018 were World Confederation for Physical Therapy Member Organisations from 39 countries.
We also thank the individual physiotherapists who have made a donation to PEDro during 2018.
But PEDro is faces significant financial challenges. We need more partners to help us continue the work we do and keep PEDro free and accessible around the world. From private practices to hospitals, government departments and universities, we can tailor a sponsorship package to suit any organisation. If your organisation would like to invest in the future of physiotherapy, please contact us.
Another way we can pay for PEDro and keep it free is through donations from users. You can choose an amount that suits your budget. We truly appreciate your help.
PEDro update (4 March 2019)
PEDro contains 42,815 records. In the 4 March 2019 update you will find:
- 33,501 reports of randomised controlled trials (32,676 of these trials have confirmed ratings of methodological quality using the PEDro scale)
- 8,642 reports of systematic reviews, and
- 672 reports of evidence-based clinical practice guidelines.
PEDro update (March 2019)
PEDro was updated on 4 March 2019. For latest guidelines, reviews and trials in physiotherapy visit Evidence in your inbox.
Next PEDro update (March 2019)
The next PEDro update is on Monday 4 March 2019.
PEDro searching has improved over time
When searching a database like PEDro, the results can only be as precise as the search terms used. The more sophisticated and specific the search is, the more relevant the resulting articles will be. For busy clinicians, using time to effectively identify relevant articles is vital. In order to improve the user experience, PEDro provides video tutorials on how to structure a search. Pop-up messages are also provided for common search errors (eg, using Boolean operators).
A recent paper analysed the content of PEDro searches to evaluate whether search quality has improved since the error messages and tutorials were implemented. Utilising data sourced from Google Analytics, the investigators compared search terms entered by users over a 6-month period in 2014-2015 to the same time period in 2017-2018. The study found a very small increase in the use of sophisticated search features (eg, truncation) and small reductions in search errors (eg, using non-ASCII characters). Overall in 2017-2018, only 6% of simple and 9% of advanced search commands used sophisticated features, while 16% of simple and 12% of advanced search commands contained errors. The content of PEDro search commands was largely similar to searches from 2014-2015. These small improvements may be due to availability of video tutorials on how to optimise searching and warnings that appear when users enter search commands containing errors. However, additional strategies to improve the quality of searches are needed.
We encourage PEDro users to think about the question they are asking before starting a search and to take advantage of the Advanced search page. Tips on formulating a clinical question are available in the PEDro “how to ask a clinical question in PICO format” tutorial. The Advanced search page includes 13 fields for entering your search terms that could help make your search more specific. For help with using the Advanced search page you can watch the “how to perform a PEDro advanced search” and “how to optimise PEDro searching” videos. These tutorials and videos are available in multiple languages on the PEDro web-site.
Zadro JR, et al. PEDro searching has improved over time: a comparison of search commands from two six-month periods three years apart. Int J Med Inform 2019;121:1-9
Systematic review found that seated exercises improve cognition in older adults with chronic health conditions
This systematic review evaluates the effect of seated exercise on impairment, activity and participation levels of older adults living with a health condition or impairment. This review included trials evaluating seated exercises of various types (eg, resistance, flexibility, range of motion, balance) in people over 65 years of age compared to other exercises or usual care. Methodological quality was evaluated with the PEDro scale, and the quality of evidence of each meta-analysis was assessed using the Grades of Research, Assessment, Development and Evaluation (GRADE) approach. Fourteen trials (n = 921 participants) were included. All outcomes classified by the International Classification of Functioning were considered for this review. The sample was predominantly composed of women. Most studies (n = 9) were considered high quality. Most trials (n = 10) were conducted in residential care facilities or day care centres. The most common intervention was progressive resistance training compared to usual care or social activities. Duration of interventions ranged from six weeks to seven months, with most spanning twelve weeks. Meta-analysis of four trials (n = 141 participants) provided low-quality evidence that seated exercise had a large positive effect on cognition when compared to usual care or social activity (standardised mean difference 1.20, 95% CI 0.25 to 2.16). Meta-analysis of three trials (n = 158 participants) provided moderate quality evidence that seated exercise, compared to social activities, did not have an effect on balance (standardised mean difference 0.13, 95% CI -0.19 to 0.44). Meta-analysis of 3 trials (n = 45 participants) provided low quality evidence that seated exercise did not have an effect on activity as assessed by Timed Up and Go Test (standardised mean difference 0.28, 95% CI -1.08 to 1.63) compared to social activities. In older adults with chronic health conditions, seated exercise was better than usual care to improve cognition, but no better than social activities in improving balance and activity.
Sexton BP, et al. To sit or not to sit? A systematic review and meta-analysis of seated exercise for older adults. Australas J Ageing 2018 Dec 13:Epub ahead of print
Read more on PEDro.