Article Figures & Data
Tables
Characteristic Phase 1 Phase 3 Potentially eligible patients Number invited (potentially eligible) 630 1,017 Age, mean (SD), y 47.9 (16.8) 49.2 (15.8) Sex, female, No. (%) 340 (54) 570 (56) Excluded, No. (%)a 28 (4) 47 (5) Nonrespondents, No. (%) 234 (37) 416 (41) Respondents, No. (%) 368 (58) 554 (55) Consented to medical record review, No. (%) 307 (83) 421 (76) Followed up at 2 mo, No. (%)b 254 (69) 332 (60) Followed up at 6 mo, No. (%)c 233 (63) 314 (57) Participating patients Age, mean (SD), y 53.0 (15.0) 54.1 (14.8) Sex, female, No. (%) 202 (55) 330 (60) Routine/manual laborer, No. (%)d 102 (35) 157 (37) Currently in paid employment, No. (%) 227 (62) 323 (59) Time off work for back pain, No. (%)e 109 (49) 133 (42) Disability: RMDQ score, mean (SD)f 8.7 (5.9) 8.4 (5.7) Pain intensity: NRS rating, mean (SD)g 5.3 (2.4) 5.0 (2.6) Duration of back pain episode, No. (%) <1 month 75 (20) 94 (17) 1–3 months 62 (17) 102 (18) 3–6 months 75 (20) 111 (20) 6 months to 3 years 82 (22) 130 (24) >3 years 74 (20) 117 (21) Leg pain, No. (%) 279 (76) 408 (74) Risk group, No. (%) Low 136 (37) 214 (39) Medium 151 (41) 232 (42) High 81 (22) 108 (20) -
NRS = numerical rating scale; RMDQ = Roland-Morris Disability Questionnaire.
-
↵a Exclusions: 8 patients in phase 1 and 23 patients in phase 3 had either died or moved primary care practice; 1 and 0 were unable to respond to questions; 11 and 19 did not want to participate; 6 and 4 did not have back pain; 1 and 0 had comorbid problems that were a priority; 1 and 1 were unable to take part.
-
↵b Losses to follow-up at 2 months: 102 patients in phase 1 and 217 patients in phase 3 did not respond to mailing; 9 and 0 had either died or moved primary care practice since baseline; 2 and 3 did not want to continue to take part; 0 and 1 did not have back pain; 1 and 1 had comorbid problems that were a priority. All baseline participants were retained for the purposes of intention-to-treat analysis (by imputation).
-
↵c Losses to follow-up at 6 months: 118 patients in phase 1 and 235 patients in phase 3 did not respond to mailing; 5 and 3 had either died or moved primary care practice since 2-month follow-up; 0 and 1 had comorbid problems that were a priority; 0 and 1 did not have back pain. All baseline participants were retained for the purposes of intention-to-treat analysis (by imputation).
-
↵d Based on major groups 5 to 9 of the UK Standard Occupation Classification (SOC, 2000) for current or most recent paid job.
-
↵e Respondents who were currently in paid employment at baseline and had time off in the 12 months before baseline.
-
↵f On a scale of 0 to 24: 0 = no disability, 24 = maximum disability.
-
↵g On a scale of 0 to 10: 0 = no pain, 10 = pain as bad as could be.
-
Outcome Phase 1
(n = 368)
No. (%)Phase 3
(n = 554)
No. (%)Odds Ratio
(95% CI)P
ValueaRisk-appropriate referral to physical therapy Not referred within the low-risk group 89 (65) 123 (68) 1.12 (0.74–1.68) .64 Referred within the medium-/high-risk groups 93 (40) 270 (72) 2.36 (1.80–3.10) <.001 Reconsulted family physician 88 (29) 159 (38) 1.51 (1.10–2.07) .01 Prescribed medicationsb Nonsteroidal anti-inflammatory drugs 136 (44) 159 (38) 0.76 (0.57–1.03) .08 Antidepressants 31 (10) 42 (10) 0.99 (0.61–1.61) .96 Nonopioids 66 (22) 53 (13) 0.53 (0.35–0.78) .001 Opioids 88 (29) 201 (48) 2.27 (1.66–3.11) <.001 Strongc 73 (24) 114 (27) 1.19 (0.85–1.67) .31 Weak only 15 (5) 87 (21) 5.07 (2.87–8.97) <.001 Neuromodulators/antiepileptics 18 (6) 20 (5) 0.80 (0.42–1.54) .51 Issued a sickness certificate 45 (15) 40 (9) 0.61 (0.39–0.96) .03 Had diagnostic tests ordered Blood test 110 (36) 154 (37) 1.03 (0.76–1.40) .84 MRI scan/radiograph 47 (15) 55 (13) 0.83 (0.55–1.27) .39 -
MRI = magnetic resonance imaging.
-
↵a Derived by χ2 test.
-
↵b Ascertained from British National Formulary codes: nonsteroidal drugs (10.1.1), antidepressants (4.3.1), nonopioids (4.7.1), opioids (4.7.2), and antiepileptics (4.8.1).
-
↵c Buprenorphine, butrans, co-codamol, codeine phosphate, DF Forte, Dtrans, fentanyl, kapake, matrifen, morphine, nabumetaone, oxycodone, oxycontin, OxyNorm, Palladone, pethidine, tramadol, trazodone, Zydol.
-
Outcome Phase 1
(n = 368)Phase 3
(n = 554)Effect Estimatea
(95% CI)P
ValueaFunction and pain Disability: change in RMDQ score 2.2 (6.0) 2.7 (5.5) 0.71 (0.06 to 1.36)b .03 Achieved MCIDc, No. (%) 154 (42) 261 (47) 1.24 (0.92 to 1.67)d .15 Pain intensity: change in NRS rating 1.7 (2.8) 1.9 (3.2) 0.29 (−0.05 to 0.63)b .09 Physical function: change in SF-12 PCS score −3.7 (11.5) −3.9 (16.3) −0.78 (−2.47 to 0.92)b .35 Psychosocial outcomes Catastrophizing: change in PCS-CAT score 1.7 (8.3) 1.8 (10.3) 0.28 (−1.13 to 1.68)b .67 Fear avoidance: change in TSK score 2.5 (8.3) 3.6 (11.2) 1.58 (0.53 to 2.62)b .006 Anxiety: change in HADS subscale score 1.0 (44) 1.2 (4.7) 0.34 (−0.29 to 0.97)b .27 Depression: change in HADS subscale score 1.0 (40) 1.4 (3.7) 0.46 (−0.07 to 0.98)b .08 Mental health: change in SF12 MCS score −1.9 (14.3) −2.1 (13.7) −0.56 (−2.77 to 1.64)b .58 Pain self-efficacy: change in PSEQ score −7.6 (12.8) −7.2 (15.5) −0.45 (−2.16 to 1.26)b .60 Global change from baseline, No. (%) Completely recovered 38 (10) 63 (11) 1.22 (0.93 to 1.59)d .16 Much better 102 (28) 178 (32) Better 88 (24) 129 (23) No change 95 (26) 127 (23) Worse/much worse 45 (12) 57 (10) Risk group, No. (%) Low 240 (65) 404 (73) 1.49 (1.05 to 2.12)d .03 Medium 96 (26) 122 (22) High 32 (9) 28 (5) Work losse Absenteeism: days off work since baseline 7.9 (23.5) 4.3 (14.5) 0.47 (0.24 to 0.92)f .03 Presenteeism: reduced productivity at work 2.2 (2.5) 2.0 (2.5) 0.17 (−0.42 to 0.75)b .57 Satisfaction with care received, No. (%)g Satisfied 161 (71) 215 (70) 1.15 (0.84 to 1.59)d .39 Neutral 23 (10) 41 (13) Not satisfied 44 (19) 52 (17) Satisfaction with results of care, No. (%)g Very satisfied/satisfied 133 (59) 209 (70) 1.50 (1.04 to 2.16)d .03 Neutral 39 (17) 45 (15) Not satisfied 52 (23) 46 (15) -
CSQ-CAT = Coping Strategies Questionnaire – CATastrophizing subscale; HADS = Hospital Anxiety and Depression Scale; MCID = minimal clinically important difference; MCS = Mental Component Subscale; PCS = Physical Component Subscale; PSEQ = Pain Self-Efficacy Questionnaire; RMDQ = Roland and Morris Disability Questionnaire; SF-12 = Short Form 12; TSK = Tampa Scale of Kinesiophobia.
-
Note: Values are mean (SD) unless otherwise noted.
-
↵a Effect estimate and 95% confidence interval derived by regression analyses adjusted for age, sex, general practice, baseline RMDQ, duration of pain, and corresponding baseline value.
-
↵b Mean difference derived by linear regression.
-
↵c Based on an MCID (≥30% reduction in RMDQ score from baseline and rating of “completely recovered” or “much better” or “better” according to the global change question).
-
↵d Odds ratio derived by binary/ordinal logistic regression.
-
↵e Based on 290 of 547 respondents who reported being currently employed at 6-month follow-up.
-
↵f Incidence rate ratio derived by robust Poisson regression.
-
↵g Based on 536 respondents for the care received question and 524 for the results of care question at the 6-month follow-up.
-
Low Risk Medium Risk High Risk Outcome Phase 1
(n = 136)Phase 3
(n = 214)P
ValueaPhase 1
(n = 151)Phase 3
(n = 232)P
ValueaPhase 1
(n = 81)Phase 3
(n = 108)P
ValueaProcess of care outcomes Reconsulted physician, No. (%) 19 (18) 33 (22) .43 39 (31) 74 (41) .056 30 (42) 52 (58) .03 Prescribed medications, No. (%) Nonsteroidal anti-inflammatory drugs 46 (43) 43 (28) .01 61 (48) 71 (39) .15 29 (40) 45 (51) .19 Antidepressants 2 (2) 8 (5) .16 19 (15) 18 (10) .20 10 (14) 16 (18) .48 Nonopioids 11 (10) 11 (7) .39 30 (23) 30 (17) .14 25 (35) 12 (13) .001 Opioids 19 (18) 47 (31) .02 44 (34) 95 (53) .001 25 (35) 59 (66) <.001 Strong opioids 12 (11) 25 (16) .24 37 (29) 49 (27) .75 24 (33) 40 (45) .13 Weak opioids 7 (7) 22 (14) .046 7 (5) 46 (26) <.001 1 (1) 19 (21) <.001 Neuromodulators/antiepileptics 0 (0) 1 (1) 1.00 10 (8) 4 (2) .02 8 (11) 15 (17) .30 Issued sickness certification, No. (%) 10 (9) 11 (7) .54 21 (16) 16 (9) .046 14 (20) 13 (15) .41 Blood test ordered, No. (%) 33 (31) 50 (33) .73 46 (36) 65 (36) .98 31 (43) 39 (44) .92 MRI scan/radiograph ordered, No. (%) 12 (11) 10 (7) .19 20 (16) 29 (16) .91 15 (21) 16 (18) .65 Clinical outcomes Disability: change in RMDQ score 0.9 (5.8) 0.9 (4.5) .87 3.4 (6.3) 3.5 (6.0) .21 2.3 (5.8) 4.8 (6.8) .004 Achieved MCID,b No. (%) 67 (49) 103 (48) .68 66 (44) 113 (49) .17 22 (27) 45 (42) .06 Pain intensity: change in NRS rating 1.0 (2.9) 0.8 (3.0) .83 2.3 (3.0) 2.4 (3.1) .52 1.9 (2.6) 2.9 (3.3) .02 Physical function: change in SF-12 PCS score −2.2 (15.2) −2.6 (16.5) .98 −5.7 (13.9) −4.0 (11.9) .79 −2.3 (13.1) −6.1 (14.8) .051 Catastophizing: change in PCS-CAT score 0.5 (7.8) 0.5 (6.4) .86 1.2 (7.3) 1.1 (10.0) .83 4.9 (8.6) 6.0 (11.7) .36 Fear avoidance: change in TSK score 2.8 (9.2) 3.1 (8.4) .42 1.7 (7.6) 3.3 (7.5) .02 3.3 (7.1) 5.3 (12.3) .09 Anxiety: change in HADS subscale score 0.7 (4.1) 0.6 (4.2) .82 0.8 (3.7) 1.0 (4.0) .09 2.1 (5.5) 2.7 (4.3) .22 Depression: change in HADS subscale score 0.4 (4.1) 0.6 (3.8) .90 1.4 (3.3) 1.4 (3.3) .45 1.2 (4.3) 2.7 (3.6) .007 Mental health: change in SF-12 MCS score −1.1 (13.4) −0.2 (14.4) .61 −1.2 (13.8) −2.0 (12.8) .23 −4.8 (17.4) −6.4 (11.7) .63 Pain self-efficacy: change in PSEQ score −4.9 (13.7) −3.1 (13.4) .31 −9.6 (16.8) −8.4 (15.2) .72 −8.4 (12.5) −12.6 (17.0) .07 Global change since baseline: much improved, No. (%) 67 (49) 116 (54) .48 53 (35) 88 (38) .11 19 (24) 41 (38) .09 Risk group: low risk, No. (%) 117 (86) 197 (92) .50 97 (64) 169 (73) .30 35 (43) 57 (53) .02 Absenteeism since baselinec 0.5 (2.3) 0.9 (3.4) .50 11.3 (26.3) 5.3 (18.7) .005 15.5 (35.5) 9.4 (16.8) .41 Presenteeismc 1.5 (2.0) 1.2 (1.5) .32 2.3 (2.3) 2.3 (2.8) .87 3.7 (3.1) 3.2 (2.8) .62 Satisfied with care receivedd 59 (75) 81 (76) .87 77 (75) 99 (70) .39 25 (54) 35 (59) .61 Satisfied with results of cared 48 (62) 81 (76) .03 63 (64) 93 (68) .50 22 (47) 35 (61) .14 -
Notes: Values are mean (SD) unless otherwise noted. See Table 3 for abbreviations.
-
↵a P values were derived by χ2 test for the process outcomes, and by regression analyses for the clinical outcomes (adjusted for age, sex, family physician practice, baseline RMDQ score, duration of pain, and corresponding baseline value—as detailed in Table 3).
-
↵b Based on an MCID (≥30% reduction in RMDQ score from baseline) and rating of “completely recovered” or “much better” or “better” according to the global change question.
-
↵c Based on 290 of 547 respondents who reported being currently employed at the 6-month follow-up.
-
↵d Based on 536 respondents for the care received question and 524 for the results of care question at the 6-month follow-up.
-
Health Care Resource Cost, Mean (SD), £ Phase 1
(n = 233)Phase 3
(n = 314)Primary care contacts Family physician: surgery 36.82 (53.5) 32.38 (50.1) Family physician: home visit 0.45 (6.9) 2.04 (20.5) Practice nurse: surgery 1.51 (6.2) 0.80 (3.8) Practice nurse: home visit 0.09 (1.3) 0.06 (1.1) Physiotherapy service NHS 33.15 (60.8) 45.09 (77.2) Private health care 13.82 (63.7) 11.90 (40.8) Hospital-based care NHS consultant 33.64 (84.5) 29.95 (80.7) NHS admissions 9.79 (91.6) 12.84 (116.9) NHS radiograph 4.39 (12.9) 3.97 (11.0) NHS CT scan 0.43 (6.6) 1.32 (11.2) NHS MRI scan 16.90 (55.0) 17.38 (53.9) NHS blood tests 0.30 (2.2) 0.06 (1.0) NHS epidural injections 3.51 (26.6) 2.09 (19.9) Private consultant 15.54 (71.5) 9.25 (55.2) Private admissions 4.97 (75.8) 3.68 (65.3) Private diagnostic tests 3.07 (23.3) 3.80 (25.4) Private epidural injections 0.88 (13.4) 0.02 (0.1) Other health care professionals NHS acupuncture 2.44 (17.5) 1.83 (20.6) NHS osteopathy 0.16 (2.5) 3.19 (20.8) NHS “other” 8.52 (46.7) 4.96 (33.9) Private acupuncture 11.90 (67.1) 3.34 (23.3) Private osteopathy 11.08 (37.0) 9.02 (38.7) Private “other” 15.41 (63.0) 4.85 (24.7) Out-of-pocket treatmentsa 15.69 (53.0) 17.50 (83.7) Prescribed medicationa 17.32 (80.1) 6.56 (19.5) Work absence costs: time off work due to low back painb 758.75 (2481.3) 358.95 (1160.4) -
CT = computed tomography; MRI = magnetic resonance imaging; NHS = National Health Service.
-
Notes: Analyses were among patients providing responses to the resource use questions at 6 months.
-
↵a Aggregate estimate that combines analgesics (nonopioid and weak opioid), nonsteroidal anti-inflammatory drugs, gels, creams, sprays, aids, and appliances.
-
↵b The estimation of indirect costs focused on respondents in paid employment at 6-month follow-up: 132 of 229 (58%) in phase 1 and 158 of 310 (51%) in phase 3.
-
Population Estimated Cost, Mean (SD), £ Mean Difference
(95% CI), £P Value Phase 1 Phase 3 Intention-to-treat (N = 922) 276.48 (585.3) 242.94 (500.5) −33.54 (−98.6 to 31.5) .31 Complete resource use data (N = 547) 261.79 (362.0) 227.90 (309.8) −33.89 (−91.8 to 24.0) .25 Complete resource use and EQ-5D data (N = 447) 287.29 (380.1) 221.25 (312.5) −66.04 (−132.2 to 0.2) .05 -
Note: The mean difference was calculated as the phase 3 estimate minus the phase 1 estimate. Confidence intervals were generated using conventional parametric methods.
-
Additional Files
Supplemental Tables 1-6 & Supplemental Figure 1
Supplemental Tables 1-6 & Figure 1
Files in this Data Supplement:
- Supplemental data: Tables & Figure - PDF file, 7 pages
The Article in Brief
Effect of Stratified Care for Low Back Pain in Family Practice (IMPaCT Back): A Prospective Population-Based Sequential Comparison
Nadine E. Foster , and colleagues
Background Back pain can be difficult to resolve and may lead to disability. This study used stratified care, in which a prognostic screening tool classifies patients by level of risk for disability and matches them with appropriate treatment. In particular, the study examines whether the use of stratified care in a primary care setting is clinically effective, leads to more targeted use of healthcare resources by changing physician referral behavior, and reduces healthcare costs.
What This Study Found Risk-stratified care for low back pain in primary care results in significant improvements in patient disability outcomes and reductions in work absence without an increase in health care costs. Patients with low back pain who received stratified care had modest improvements in physical function, fear avoidance beliefs, satisfaction with care and time off work. Average time off work was 50% shorter (4 vs 8 days) and the proportion of patients given sickness certifications was 30% lower (9% vs 15%) in the post-intervention group. Significant changes to physician clinical behavior included increased numbers of risk-appropriate referrals to physical therapy, reduced prescribing of nonsteroidal medications and fewer sickness certifications. There was also a small overall reduction in health care resource use and large societal cost savings due to fewer periods of pain-related work absence.
Implications
- The authors recommend widespread implementation of stratified care based on its association with benefits for patients and more targeted use of health care resources without increasing health care costs.
In this issue
Jump to section
Related Articles
Cited By...
- Evaluating the STarTBack stratified treatment approach for low back pain: exploring study-level factors potentially explaining differences in results of studies - a literature review
- Low back pain management in primary healthcare: findings from a scoping review on models of care
- A Stratified Approach for Managing Patients With Low Back Pain in Primary Care (SPLIT Program): A Before-and-After Study
- Unsupervised subgrouping of chronic low back pain patients treated in a specialty clinic
- Patients and physiotherapists perspectives on implementing a tailored stratified treatment approach for low back pain in Nigeria: a qualitative study
- Study protocol for the development and internal validation of Schizophrenia Prediction of Resistance to Treatment (SPIRIT): a clinical tool for predicting risk of treatment resistance to antipsychotics in first-episode schizophrenia
- Study protocol for the development and internal validation of SPIRIT (Schizophrenia Prediction of Resistance to Treatment): A clinical tool for predicting risk of treatment resistance to anti-psychotics in First Episode Schizophrenia
- Feasibility of delivering and evaluating stratified care integrated with telehealth ('Rapid Stratified Telehealth) for patients with low back pain: protocol for a feasibility and pilot randomised controlled trial
- Implementation of a novel stratified PAthway of CarE for common musculoskeletal (MSK) conditions in primary care: protocol for a multicentre pragmatic randomised controlled trial (the PACE MSK trial)
- Feasibility study of a multicentre cluster randomised control trial to investigate the clinical and cost-effectiveness of a structured diagnostic pathway in primary care for chronic breathlessness: protocol paper
- Which Exercise for Low Back Pain? (WELBack) trial predicting response to exercise treatments for patients with low back pain: a validation randomised controlled trial protocol
- Identifying psychosocial characteristics that predict outcome to the UPLIFT programme for people with persistent back pain: protocol for a prospective cohort study
- Do sensorimotor cortex activity, an individuals capacity for neuroplasticity, and psychological features during an episode of acute low back pain predict outcome at 6 months: a protocol for an Australian, multisite prospective, longitudinal cohort study
- Can the STarT Back Tool predict health-related quality of life and work ability after an acute/subacute episode with back or neck pain? A psychometric validation study in primary care
- Beliefs, Behavior, and Back Pain: Insights into Psychologically Informed Management
- Differences among primary care patients with different mechanical patterns of low back pain: a cross-sectional investigation
- Prediction of prognosis for people off sick with upper extremity musculoskeletal disorders
- Does a modified STarT Back Tool predict outcome with a broader group of musculoskeletal patients than back pain? A secondary analysis of cohort data
- Preliminary Evaluation of a Modified STarT Back Screening Tool Across Different Musculoskeletal Pain Conditions
- Pragmatic Implementation of a Stratified Primary Care Model for Low Back Pain Management in Outpatient Physical Therapy Settings: Two-Phase, Sequential Preliminary Study
- Staged Approach for Rehabilitation Classification: Shoulder Disorders (STAR-Shoulder)
- In This Issue: Building Blocks for Improving Practice
- Physical Therapy for Low Back Pain: What Is It, and When Do We Offer It to Patients?