A User-Driven Web Application to Explore Treatment Options for Lower Back Pain
|
If you are the presenter of this abstract (or if you cite this abstract in a talk or on a poster), please show the QR code in your slide or poster (QR code contains this URL). |
Abstract
Background
Chronic lower-back pain (CLBP) affects 30 million Americans yearly, at an annual cost of over $100 billion. Few patients find complete pain relief in a sea of uncertain treatment alternatives. For physicians CLBP involves uncertainty over symptoms and patients’ preconceived notions. These mutual frustrations can lead to an antagonistic, rather than collaborative, approach. In addition, a user-driven approach to treatment decisions is time-consuming and difficult to achieve in an office visit. Thus, there is a need for an application for knowledge-translation and decision-making in CLBP that facilitates interactive information sharing and incorporates patients’ demographics, values, and preferences. Our Web 2.0 application collects information from different sources (clinical databases, users themselves), combines and weights the information based on the users' inputs and preferences, and presents it to users in a meaningful way. It also has an asymmetrical relationship advantage (leveraging the same information platform with two different entry points--lay and expert), which is a core Web 2.0 pattern. Analytical Hierarchy Processing (AHP) is a framework for structuring and evaluating competing alternatives that has been used in non-medical decision-making for 30 years. Through binary prompts, AHP determines relative preferences amongst complex problems. We propose a methodology for a Web-based version of AHP for CLBP.
The objective of this study was to assess the value of a Web-based Analytical Hierarchy Processing framework as a participatory, user-driven decision-support model in chronic lower-back pain.
Methods
We redesigned AHP for CLBP by incorporating the Oswestry Disability Index (ODI), a gold-standard back-pain outcome measure. ODI rates back pain on a scale of 1-100 in 10 life-areas, ordinarily ranking the value of improvement within each area on the same scale (0=pain-free). For a training dataset, the raw data (N=349) from a 2005 BMJ study that compared rehabilitation and surgery was characterized for AHP. Exploratory data analysis was conducted for each ODI life-area by treatment, including a probability distribution of each pain-state and life-area. Significant differences between surgery and rehabilitation were further examined over baseline, 12 months, and 24 months using linear regression/ANOVA. Data visualization and local polynomial regression fitting elucidated distinctions between treatment effects. Results
Data-checking found 33 patients crossed-over from rehab to surgery mid-trial. Due to uncertainty of timeframe and large effect on outcomes, they were analyzed separately. Apparent differences in baseline ODI scores between treatments disappeared over time. Surgery averaged a decrease of 10.99 from baseline to 12 months while rehab averaged a decrease of 5.93. ANOVA was significant (p=0.026). There was no significant difference between the two groups by 24 months, suggesting that surgery patients began worse than rehab but ended with approximately equal scores. Data visualization elicited a “gray area†that suggests non-linearity in the responses where surgery outperformed rehabilitation in ODI score improvement. This distinction was not statistically significant but would bear revisiting with more data. The framework was able to compute patients’ probability states for all pain levels and ODI questions. It differentiated which treatment would be better dependent on the patient’s starting level of pain and preferences. The AHP model also detected ranges for appropriate expectations of pain improvement.
Conclusions
Combining a proven decision framework with the Web 2.0 values of information sharing and user-centered content, AHP holds the potential to improve CLBP decision-making for patients and practitioners. It may also level patient expectations and reduce regret. We have begun developing a systematic methodology to apply AHP to other diseases.
Chronic lower-back pain (CLBP) affects 30 million Americans yearly, at an annual cost of over $100 billion. Few patients find complete pain relief in a sea of uncertain treatment alternatives. For physicians CLBP involves uncertainty over symptoms and patients’ preconceived notions. These mutual frustrations can lead to an antagonistic, rather than collaborative, approach. In addition, a user-driven approach to treatment decisions is time-consuming and difficult to achieve in an office visit. Thus, there is a need for an application for knowledge-translation and decision-making in CLBP that facilitates interactive information sharing and incorporates patients’ demographics, values, and preferences. Our Web 2.0 application collects information from different sources (clinical databases, users themselves), combines and weights the information based on the users' inputs and preferences, and presents it to users in a meaningful way. It also has an asymmetrical relationship advantage (leveraging the same information platform with two different entry points--lay and expert), which is a core Web 2.0 pattern. Analytical Hierarchy Processing (AHP) is a framework for structuring and evaluating competing alternatives that has been used in non-medical decision-making for 30 years. Through binary prompts, AHP determines relative preferences amongst complex problems. We propose a methodology for a Web-based version of AHP for CLBP.
The objective of this study was to assess the value of a Web-based Analytical Hierarchy Processing framework as a participatory, user-driven decision-support model in chronic lower-back pain.
Methods
We redesigned AHP for CLBP by incorporating the Oswestry Disability Index (ODI), a gold-standard back-pain outcome measure. ODI rates back pain on a scale of 1-100 in 10 life-areas, ordinarily ranking the value of improvement within each area on the same scale (0=pain-free). For a training dataset, the raw data (N=349) from a 2005 BMJ study that compared rehabilitation and surgery was characterized for AHP. Exploratory data analysis was conducted for each ODI life-area by treatment, including a probability distribution of each pain-state and life-area. Significant differences between surgery and rehabilitation were further examined over baseline, 12 months, and 24 months using linear regression/ANOVA. Data visualization and local polynomial regression fitting elucidated distinctions between treatment effects. Results
Data-checking found 33 patients crossed-over from rehab to surgery mid-trial. Due to uncertainty of timeframe and large effect on outcomes, they were analyzed separately. Apparent differences in baseline ODI scores between treatments disappeared over time. Surgery averaged a decrease of 10.99 from baseline to 12 months while rehab averaged a decrease of 5.93. ANOVA was significant (p=0.026). There was no significant difference between the two groups by 24 months, suggesting that surgery patients began worse than rehab but ended with approximately equal scores. Data visualization elicited a “gray area†that suggests non-linearity in the responses where surgery outperformed rehabilitation in ODI score improvement. This distinction was not statistically significant but would bear revisiting with more data. The framework was able to compute patients’ probability states for all pain levels and ODI questions. It differentiated which treatment would be better dependent on the patient’s starting level of pain and preferences. The AHP model also detected ranges for appropriate expectations of pain improvement.
Conclusions
Combining a proven decision framework with the Web 2.0 values of information sharing and user-centered content, AHP holds the potential to improve CLBP decision-making for patients and practitioners. It may also level patient expectations and reduce regret. We have begun developing a systematic methodology to apply AHP to other diseases.
Medicine 2.0® is happy to support and promote other conferences and workshops in this area. Contact us to produce, disseminate and promote your conference or workshop under this label and in this event series. In addition, we are always looking for hosts of future World Congresses. Medicine 2.0® is a registered trademark of JMIR Publications Inc., the leading academic ehealth publisher.
This work is licensed under a Creative Commons Attribution 3.0 License.