As rehabilitation professionals are pressured in the value based care model, more and more attention will be taken to guesstimate whether a patient will or will not improve with services.
This study looked at a variety of factors with the ultimate goal of predicting if the patient was a low risk, normal risk or high risk with regard to improving with physical therapy services. (I view medium risk as normal risk.) For some reason, I am immediately thinking of the Pareto principle (the 80/20 rule). Granted, as I look at Figure 1 in the full text, in this case the 80/20 rule is wrong. It appears that 3% were low risk, 89.8% were normal risk and 7.2% were high risk. It will be important to readily identify patients at a high risk of not really improving in a reasonable amount of visits.
Risk stratification is a generic type of prediction. Its focus is to predict a level of risk. Risk stratification isn't designed to provide specific details about the outcome of care.
What I found interesting is thinking about risk stratification and risk adjustment. If I were to use my own terms, risk stratification is more like providing the likelihood of improving. When I think of risk adjustment, the goal is to have a more precise prediction about the outcome of care.
The same factors that were highlighted in the abstract that determined low, normal and high risk are also included in FOTO's risk adjustment process. FOTO's risk adjustment process includes far more variables than those mentioned in the article. Last month FOTO improved its risk adjustment process to even more accurately predict functional change, number of visits and duration of care. The predicted outcome is adjusted based on the impact each patient factor has on the outcome of care. Even though FOTO doesn't provide generic low, normal or high risk information, the patient specific report can easily indicate if the person is a high risk. The high risk patients will need more visits and have little improvement predicted.
Below you will find a quick view of the abstract.
Musculoskeletal shoulder pain is commonly treated in physical therapy. There is inconsistency in the literature regarding patient characteristics related to prognosis. Having prognostic information could be useful for improving clinical efficiency and decreasing the cost of associated care. The objective of this study was to identify predictive characteristics related to patients with shoulder pain who have a high-risk of a bad prognosis (lowest functional recovery compared with visit utilization) as well as those who are at low-risk of a bad prognosis (highest functional recovery compared with visit utilization).
We completed a secondary analysis of a retrospective cohort using data obtained from an existing commercial outcomes database. Data from 5214 patients with shoulder pain were analysed to determine predictive characteristics that identify patients who either have a low-risk or a high-risk of a bad prognosis to physical therapy care. Multinomial regression was used to identify significant patient characteristics predictive of treatment response.
Statistically significant predictors of high-risk categorization included older age, no surgical history, insurance designated as worker's compensation, litigation or automotive and three or more co-morbidities. Predictors of low risk categorization were younger age, shorter duration of symptoms, no surgical history and payer type.
Selected variables were associated with both poor and good recovery. Further research on prognosis, efficacy of physical therapy care and cost appear warranted for patients with shoulder pain.
J Eval Clin Pract. 2017 Apr;23(2):257-263. doi: 10.1111/jep.12591. Epub 2016 Jun 29.