Background
The “Subgroups for Targeted Treatment” (STarT) risk stratification approach
Will the STarT back risk stratification strategy work in the US?
Methods/design
Development and implementation of the QI strategy
Assembled the project team
Elicited perspectives of key stakeholders
Identified treatment options for each risk subgroup
-
Low Risk (~40 % of patients): The PCP can manage most patients in this category through minimal interventions without referral to an additional provider [30]. For these patients the provider should conduct a brief assessment to rule out potentially serious causes of back pain (i.e., “red flags”), elicit and listen to patients’ concerns, provide reassurance about the positive prognosis and self-care recommendations to relieve pain (i.e., appropriate physical activity, use of pain medications and avoiding bed rest). PCPs were trained and encouraged to recommend that their patients access online DVDs that reinforced information about acute or chronic back pain and the importance of self-care.
-
Moderate Risk (~40 % of patients): Patients in this group should be offered additional guideline-recommended treatment options, particularly those involving exercise and activating treatments that could reduce fear of movement (i.e., PT and yoga). Patients not interested in activating treatments should be offered the more passive options (acupuncture, spinal manipulation, or massage therapy) in the hope these treatments will help decrease their pain and prepare them for transition to more active approaches, including the Living Well with Chronic Conditions self-management classes available at GH.
-
High Risk (~20 % of patients): High risk patients are on average more complex (distressed and disabled) than patients in the low or medium risk subgroup, therefore, in addition to the treatment options for moderate risk patients, the best available treatment options for patients in this subgroup at GH was psychologically-informed practice delivered by PTs specially-trained for this initiative or referral to a psychologist for CBT. “Psychologically informed practice offers a systematic approach to the integration of physical and psychological approaches to treatment for the management of people with low back pain…” [19]. Here, ‘psychological’ refers to the beliefs/expectations, emotional and behavioral responses associated with low back. Unfortunately, access CBT from a psychologist was very limited. PCPs were also encouraged to proactively follow-up with high risk patients within 2 weeks.
Incorporated STarT back tool, recommended treatments, and decision aids into the EHR
Developed training modules for primary care teams
Overview | |
---|---|
Session 1 |
Introduction
Physician leader of the back pain care quality improvement project introduced the researchers, the rationale and importance of reorganizing care for back pain and overview of the project aims and activities. Focused on getting clinicians to support the project and excited about the opportunity to improve back pain care in their clinics. Introduced the STarT Back tool and risk stratification strategy. |
Session 2 |
Using the STarT Back Tool
Focused on getting PCPs and staff comfortable with administering the STarT Back approach, including use in the EHR (EPIC), scoring of tool, and understanding matched treatment recommendations for each risk level. Discussions of how to find and use the GH Back Pain Guidelines, attach the STarT Back tool in secure messages to patients, enter patient responses into the EHR, view results, use tools within the EHR to enhance visits for back pain such as patient instructions and ordering patient centered back pain (acute and chronic) DVDs. |
Session 3 |
Improving Diagnosis and Ruling out Red Flags
Check-in on clinic’s use of the STarT Back tool. Used patient case examples to provide a “refresher” on how to conduct differential diagnoses of common back pain problems (focused on lower back, L2-L4, L5, S1). Reviews included how to conduct an efficient exam, common errors in examinations, appropriate use and interpretation of diagnostic imaging, red flags for serious conditions, how to communicate to patients during the examination. |
Session 4 |
Talking with Patients about Chronic Pain
Training focused on ways to communicate more effectively with patients about chronic pain including: 1) preferred language in discussing pain, 2) ways to better communicate anatomical links to pain, 3) explaining what chronic pain is (pain centralization, gate theory, reoccurrence of pain – having continued pain with no injury), 4) talking about red flags and when to return to primary care, 5) focusing on improving function rather than reducing pain, and 6) how to discuss outcomes from the STarT Back tool and shared decision making around treatment options. |
Session 5 |
Improving Partnership between PCPs and PTs
This session brought together PCPs and PTs for an interactive discussion on how to improve team based care for patients. Topics included: how PT was providing improved care based on training, shared responsibility and roles of providers, how to integrate the STarT Back tool across the departments, and how to collaboratively work together to help patients not showing improvement. |
Session 6 |
Complementary and Alternative Medicine Treatments for Chronic Back Pain
Focused on building an understanding of the role that CAM providers (acupuncturists, chiropractors, massage therapists) and yoga classes can play for back pain patients. Emphasized practical information: brief description of each CAM modality, scientific evidence for their effectiveness, contraindications, dosing. CAM therapies referrals were linked to STarT Back tool risk category, emphasized the use of active over passive therapies, and how to conduct referrals within the healthcare system. |
Training for physical therapists
Planned Topics | |
---|---|
Day 1 | • Description of STarT Back Trial, IMPACT study and other related research • Description of stratified care • Myths and facts about patients that have chronic pain • Research on pain models, the complexity of pain experiences, with special emphasis on moving away from seeing pain as an indication of tissue damage. • Research on neurophysiology of pain |
Day 2 | • Research on neurophysiology of pain (continued) • Review of key factors that contribute to development and maintenance of pain related disability • Communication skills for working with patients with disabling chronic pain |
Day 3 | • Assessment of high risk patients • Managing/treating high risk patients • Integrating the psychosocial approach into manual therapy |
Day 4 | • Explaining pain • Managing expectations • Facilitating behavioral change/goal setting |
Day 5 | • Managing disability • Vocational rehabilitation • Clinical decision making and treatment planning • Monitoring and modifying treatment plans |
Implemented risk-stratification strategy
Protocol for evaluating the effects of implementing the STarT back strategy
Effect on patient outcomes
Patient recruitment
Patient outcome measures (Table 3)
Measures | Baseline | 2-month | 6-month |
---|---|---|---|
Baseline Characteristics | |||
Patient characteristics: education, income, employment). (Age, gender, race, and ethnicity extracted from EHR) | x | ||
Back pain problem (duration, pain elsewhere) | x | ||
Risk of Poor Outcome (STarT Back tool) | x | x | x |
Primary Outcomes | |||
Back pain-related physical function (RMDQ) | x | x | x |
Back pain severity (0–10 scale) | x | x | x |
Secondary Outcomes | |||
Depression (PHQ-8) | x | x | x |
Anxiety (GAD-7) | x | x | x |
Fear of movement (TKS-10 item version) | x | x | x |
Global improvement (PGIC) | x | x | |
Self-efficacy (PSEQ) | x | x | x |
Patient satisfaction with caring, information, treatment effectiveness | x | x | x |
Work loss in past 7 days due to LBP (hours); Effect on work productivity (0 to 10 scale) (2 items from WPAI) | x | x | x |
Treatment-Related Information | |||
Helpfulness of treatments recommended by PCPs | x | x | |
Medications used in past week; change in medication use over time | x | x | x |
Out-of-pocket expenses | x | x |
Primary outcomes
Secondary outcomes
-
o
Fear of movement was measured by an abbreviated version of the Tampa Scale for Kinesiophobia, a validated 17-item questionnaire that quantifies excessive fear of (re)injury due to movement in pain patients [40]. The 10-item version was developed for another trial [41] had a measure of internal consistency of 0.66 [Michael von Korff, personal communication, May 1, 2016]. The total score of the 10 items was adjusted to yield a total score comparable to the 17-item version by multiplying the average item score for the items answered by 17. -
o
Global improvement with treatment. We used the Patient Global Impression of Change (PGIC) scale [42], a single question asking participants to rate improvement with treatment on a seven-point scale ranging from “very much improved” to “very much worse,” with “no change” as the mid-point. -
o
Self-efficacy was measured using the 10-item Pain Self Efficacy Questionnaire (PSEQ). It shows good internal consistency and construct validity [43]. -
o
Satisfaction with care was measured using a 10-item instrument that has been validated and able to distinguish among three dimensions of satisfaction (caring, information and treatment effectiveness) [44]. -
o
The Work Productivity and Activity Impairment (WPAI) is a self-administered instrument used to evaluate the impact of back pain on productivity. It assesses time missed from work (absenteeism), impairment while working/reduced on-the-job effectiveness (presenteeism), and overall work productivity loss (absenteeism v. presenteeism). It has been found reliable, valid, and responsive to change for several medical conditions [45]. -
o
Use of back-related medications and exercise in the past week and treatments (e.g., yoga) paid for out of pocket (which are not captured in the GH database).
Statistical analysis plan and sample size calculations
Statistical analysis plan
Period of time | Usual Care (UC) Clinics | Risk Stratification (RS) Clinics | Column Difference |
---|---|---|---|
Pre-implementation* of risk stratification (RS) | β0: Change score in UC clinics pre-implementation of RS | β0+ β1: Change score in RS clinics pre-implementation of RS | β1: Difference in change scores due to RS in the pre-implementation period. |
Post-implementation of risk stratification (RS) | β0+ β2: Change score in UC clinics post-implementation of RS | β0 + β1 + β2 + β3: Change score in RS clinics post-implementation of RS | β1 + β3: Difference in change scores due to RS in the post-implementation period. |
Row Difference | β2: Difference in change scores due solely to time (implementation period). | β2 + β3: Difference in change scores due to time and RS. | β3: Difference in change scores due solely to the RS implementation. |