- Email: [email protected]
or Neck Pain in the Netherlands Benefit From Stratification for Risk Groups According to the STarT Back Tool Classification?
Accepted Manuscript Can primary care for back and/or neck pain in the Netherlands benefit from stratification for risk groups according to the STarT Back Tool-classification? Jasper D. Bier, MSc, Janneke J.W. Sandee-Geurts, MSc, Raymond W.J.G. Ostelo, Prof, Bart W. Koes, Prof, Arianne P. Verhagen, PhD PII:
S0003-9993(17)30456-2
DOI:
10.1016/j.apmr.2017.06.011
Reference:
YAPMR 56947
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 14 November 2016 Revised Date:
31 May 2017
Accepted Date: 5 June 2017
Please cite this article as: Bier JD, Sandee-Geurts JJW, Ostelo RWJG, Koes BW, Verhagen AP, Can primary care for back and/or neck pain in the Netherlands benefit from stratification for risk groups according to the STarT Back Tool-classification?, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2017), doi: 10.1016/j.apmr.2017.06.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Primary care for back and/or neck pain
ACCEPTED MANUSCRIPT
Can primary care for back and/or neck pain in the Netherlands benefit from stratification for risk groups according to the STarT Back Tool-classification?
RI PT
Authors: 1, 2
Jasper D. Bier, MSc
Janneke J.W. Sandee-Geurts, MSc
Bart W. Koes, Prof
5,6
SC
Raymond W. J. G. Ostelo, Prof
3,4
1
1
M AN U
Arianne P. Verhagen, PhD
Department of General Practice, Erasmus MC, PO Box 2040, 3000 CA, Rotterdam, The Netherlands
2.
FS Fysio, Capelle aan den IJssel, The Netherlands
3.
Faculty of Human Movement Sciences, VU, Amsterdam, The Netherlands
4.
Fysiotherapie Kapellaan / Ouwerkerk, Vught, The Netherlands
5.
Department of Epidemiology and Biostatistics, VU University Medical Centre Amsterdam and the
TE D
1.
EMGO Institute for Health and Care Research, The Netherlands 6.
Department of Health Sciences, Faculty of Earth and Life Sciences, VU University Amsterdam, The
Correspondence to:
EP
Netherlands.
AC C
Jasper D. Bier, Department of General Practice, Erasmus MC, PO Box 2040, 3000CA Rotterdam, The Netherlands, T +31 6 26536787
E [email protected]
URL www.erasmusmc.nl
The study was approved by the medical ethics committee of the Erasmus University, Rotterdam, The Netherlands. (METC-2014-256)
The authors declare that there is no conflict of interests regarding the publication of this paper. This study was undertaken with financial support of CZ healthcare insurance company and Dutch Arthritis Association.
Primary care for back and/or neck pain
ACCEPTED MANUSCRIPT
The authors would like to thank all of the GP’s and PT’s that included patients for this study. A special thank goes out to Nynke Wildervanck for her contribution and to Steven Constandse, Guido Iken, Joost van
AC C
EP
TE D
M AN U
SC
RI PT
Broekhoven and Frans van der Kooij for their extra efforts to reach the needed sample size.
ACCEPTED MANUSCRIPT 1
Titel:
2
Can primary care for back and/or neck pain in the Netherlands benefit from stratification for
3
risk groups according to the STarT Back Tool-classification?
4 Abstract
6
Objective: To evaluate whether current Dutch primary-care clinicians offer tailored treatment to patients with
7
lower-back pain (LBP) or neck pain (NP) according to their risk stratification, based on the Keele STarT
8
(Subgroup Targeted Treatment) Back-Screening Tool (SBT).
9
Design: Prospective cohort study with 3 month follow-up
SC
RI PT
5
Setting: Primary care
11
Participants: General practitioners (GPs) and physiotherapists (PTs) included patients with non-specific LBP
12
and/or NP.
13
Interventions: Patients completed a baseline questionnaire, including the Dutch SBT, for either LBP or NP. A
14
follow-up measurement was conducted after 3 months to determine recovery (using the Global Perceived
15
Effect (GPE) scale), pain (using the Numeric Pain-Rating Scale (NPRS)) and function (using the Roland Disability
16
Questionnaire (RDQ) or the Neck Disability Index (NDI)). A questionnaire was sent to the GPs and PTs to
17
evaluate the provided treatment.
18
Main outcome measures: Prevalence of patients’ risk profile and clinicians’ applied care, and the percentage
19
of patients with persisting disability at follow-up. A distinction was made between patients receiving the
20
advised treatment and those receiving the non-advised treatment.
21
Results: In total, 12 GPs and 33 PTs included patients. After 3 months, we analyzed 184 patients with LBP and
22
100 patients with NP. In the LBP group, 52.2% of the patients were at low risk for persisting disability, 38.0%
23
were at medium risk and 9.8% were at high risk. Overall, 24.5% of the LBP patients received a low-risk
24
treatment approach, 73.5% a medium-risk and 2.0% a high-risk treatment approach. The specific agreement
25
between the risk profile and the received treatment for patients with LBP was poor for the low-risk and high-
26
risk patients (respectively 21.1% and 10.0%), and fair for medium-risk patients (51.4%). In the NP group, 58.0%
27
of the patients were at low risk for persisting disability, 37.0% were at medium risk and 5.0% were at high-risk.
28
Only 6.1% of the patients with NP received the low-risk treatment approach. The medium-risk treatment
AC C
EP
TE D
M AN U
10
ACCEPTED MANUSCRIPT approach was offered the most (90.8%) and the high-risk approach was applied in only 3.1% of the patients.
30
The specific agreement between the risk profile and received treatment for NP patients was poor for low-risk
31
and medium-risk patients (resp. 6.3% and 48.0%); agreement for high-risk patients could not be calculated.
32
Conclusion: Current Dutch primary care for patients with non-specific LBP and/or NP does not correspond to
33
the advised stratified-care approach based on the SBT as the majority of patients receive medium risk
34
treatment. The majority of “low-risk” patients are over-treated and the majority of “high-risk” patients are
35
undertreated. Although the stratified-care approach has not yet been validated in Dutch primary care, these
36
results indicate that there may be substantial room for improvement.
37
Keywords: Usual care, general practitioner, physical therapy, lower-back pain, neck pain, STarT Back Tool
38
Abbreviations:
39
General Practitioner (GP)
40
Global Perceived Effect (GPE)
41
Lower-back pain (LBP)
42
Neck Disability Index (NDI)
43
Neck Pain (NP)
44
Numeric Pain-Rating Scale (NPRS)
45
Roland Disability Questionnaire (RDQ)
46
STarT Back-Screening Tool (SBT)
47
Subgroup Targeted Treatment (STarT)
AC C
EP
TE D
M AN U
SC
RI PT
29
ACCEPTED MANUSCRIPT 48
Introduction
49
Lower-back pain and NP are major public-health problems; they are the primary and fourth causes,
50
respectively, of disability worldwide. With regard to health-seeking behavior for back and neck pain,
51
approximately 55% of patients seek healthcare, and between 12% and 32% visit a GP.
52
Non-specific LBP and NP are the main focus of most primary-care guidelines.
53
Dutch General-Practitioners Society and the Royal Dutch Physiotherapists Society divide patients with non-
54
specific LBP and NP into roughly two subgroups: 1) normal recovery, which is defined as a decrease in pain and
55
function limitation prior to consultation, and 2) a (suspected) delayed recovery.
56
The SBT is a tool, developed in England, to allocate primary-care patients with LBP to three prognostic
57
subgroups: patients at either low, medium or high risk for persisting disability. This allocation aims to ensure
58
that the appropriate stratified care is applied to patients “at risk” for persistent LBP in order to prevent it.
59
This tool’s questions are based on known negative-prognostic factors that can be influenced by treatment.
60
Furthermore, the SBT has been found to be a valid and reliable tool for subgrouping patients in the UK , and it
61
has been translated and validated in several languages, including Dutch , since its initial English publication in
62
2008.
63
corresponding targeted treatment, has yet to be implemented in the Netherlands.
64
Advised care for normal recovery is to reassure the patient and inform him or her of the positive prognosis,
65
and to advise the patient to stay active, which might be supported with the prescription of pain medication. If
66
the pain persists or worsens, the GP can refer physiotherapy to the patient. Where current guidelines advise a
67
stepped approach for patients with a delayed recovery, the SBT advises a stratified-care approach; the
68
difference between these two approaches relates to the timing of (effective) interventions. Implementing the
69
SBT in the UK has led to higher quality-adjusted life years for patients and lower health costs, and superior
70
outcomes in the high-risk group using stratified care.
71
patients with NP; however, we expect that it will not differ from the approach for LBP. Before advising on
72
implementing the SBT in the Netherlands, we need more insight into the current usual care because
73
implementation of the SBT might be unnecessary if usual care provides comparable or better outcomes.
74
Therefore, the aim of this study is to evaluate current Dutch primary care for spinal pain and whether it
75
corresponds to the advised stratified-care approach based on the SBT.
76
1
The guidelines from both the
RI PT
7,8
2–4
SC
7,9,10
M AN U
11
12
11
20
21
The Dutch version of the SBT has been modified to fit patients with NP. The SBT, and
AC C
EP
TE D
13–19
22,23
No studies have been published on stratified care for
ACCEPTED MANUSCRIPT Method
78
Design
79
The prevalence of risk groups in neck- and back-pain patients according to the SBT (PRINS) study is a
80
prospective cohort study, which includes patients with LBP or NP of any duration that consulted a GP or PT.
81
The study was approved by the medical ethics committee of the Erasmus University, Rotterdam, the
82
Netherlands (METC-2014-256).
RI PT
77
83 Participants
85
Care providers
86
We asked GPs and PTs, who work in the primary-care sector and had displayed interest in the SBT during pilot
87
projects, to participate in the PRINS study. Information about the study protocol was provided through several
88
meetings, by phone or by digital/paper documentation, and participating GPs and PTs received the study
89
protocol and a folder with patient information and informed consent forms. The majority of these GPs and PTs
90
work in small clinics in the region of Rotterdam (a maximum of 50 kilometers), the Netherlands.
M AN U
SC
84
91 Patients
93
The inclusion period for patients started in November 2014 and continued through to May 2015, and patients
94
consulting their PT or GP for LBP or NP during that period were invited to participate. Other inclusion criteria
95
were that the patient had to be 18 years old or over, could speak and read Dutch and had an email address.
96
Patients were excluded if, during the consultation, the GP or PT found red flags indicating a possible serious
97
underlying pathology (for example, an infection, a fracture, cauda equina or a tumor) responsible for the LBP
98
or NP.
99
Patients were provided with oral and written information about the aim of the study and the procedure of
100
data collection, and each patient signed an informed consent, which was handed back to the PT or GP who
101
subsequently registered the patient online. The patient immediately received an email with a link to the
102
baseline questionnaire, and if necessary, a reminder to complete the questionnaire was sent after a few days.
103
Patients who did not complete the baseline questionnaire within 7 days were excluded from the cohort.
EP
AC C
104 105
TE D
92
Treatment
ACCEPTED MANUSCRIPT 106
The patients received usual care from their GPs or PTs, who were kept blind for the results of the patients’
107
baseline questionnaire.
108 Baseline measurements
110
The baseline questionnaire consisted of questions on demographic data, the SBT (either the back or neck
111
version) and the NPRS to assess pain; the NDI or the RDQ
112
kinesiophobia; and the pain-catastrophizing scale and the EQ-5d to assess quality of life. The NPRS options
113
range from 0 (no pain) to 10 (worst imaginable pain), the RDQ consists of 24 statements with a “yes” or “no”
114
answer option and a total score ranging from 0 to 24, and the NDI consists of 10 statements with a 6-point
115
scale ranging from 0 (not limited) to 6 (completely limited) and a total score ranging from 0 to 50.
24
25
28
to assess disability; the Tampa scale for 30
SC
29
26,27
RI PT
109
M AN U
116 Outcome measure
118
Three months after inclusion, the patients received a follow-up questionnaire to assess pain (NPRS), disability
119
(RDQ or NDI) and recovery, using the GPE. The answer options on the GPE range from 1 (fully recovered) to 7
120
(worse than ever).
121
Persisting LBP disability was defined as an RDQ score ≥ 7, based on the mean of the baseline score as used by
122
Hill at al. , while persisting NP disability was defined as an NDI score ≥ 13, based on the median baseline score
123
in this cohort.
124
Three months after inclusion, the GP was sent a questionnaire to inquire about the number of visits,
125
prescribed medication, referrals to PTs or medical professionals and requested imaging and blood tests. The PT
126
received a questionnaire to inquire about treatment data such as the number and period of treatments, and
127
the aim and means of treatment. The investigator sent and received all questionnaires digitally, and they were
128
handled and stored through LimeSurvey 2.05.
EP
11
AC C
129
TE D
117
130
Analysis
131
Treatment categorization
132
The two authors utilized the following criteria to independently categorize the treatments applied by the GPs
133
and/or PTs (in case of differences between the authors, the categories were determined by consensus):
ACCEPTED MANUSCRIPT 134
•
Low-risk approach – the GP provided information, advice and some analgesics or one or two PT
135
consultations, and the treatment was hands-off and consisted of offering information, advice and
136
exercises. •
Medium-risk approach – in addition to the low-risk approach, the GP referred the patient to a PT, and
138 139
the PT performed an evidence-based intervention. •
RI PT
137
High-risk approach – in addition to the medium-risk approach, the GP referred the patient to either a
140
PT specialized in treating patients with a psychosomatic approach, a psychologist or equivalent, and
141
the PT assessed bio-psychosocial risk factors and used cognitive behavioral principles as
142
interventions.
SC
31
143 Statistical analysis
145
We described the characteristics of the clinicians and patient population using frequencies (means with
146
standard deviations). Next, we calculated the frequencies of patients per SBT-risk profile and the clinicians’
147
treatment approaches. The specific agreement calculates percentages of advised-treatment approaches for
148
each risk profile separately. For example, patients who were “low-risk” at baseline and treated as such are
149
calculated as a proportion of patients that were “low-risk” on either of the two measurements. We modified
150
the specific agreement to fit a 3x3 table, as illustrated in Table 1, because the original method is done in a 2x2
151
table. We rated a specific agreement < 40% as poor, 40% to 59% as fair, 60% to 74% as good and 75% to 100%
152
as excellent. Lastly, we calculated the percentage of patients with persisting disability per risk profile; this was
153
done for the groups receiving the advised and non-advised treatments.
EP
TE D
32
AC C
154
M AN U
144
155
Results
156
Study population
157
The GPs and PTs originally included 370 patients; however, 86 patients failed to fill in the baseline
158
questionnaire and were excluded. The characteristics of the 284 remaining patients are presented in Table 2.
159
The 12 GPs included 103 patients (26 with NP and 77 with LBP) and the 33 PTs included 181 patients (74 with
160
NP and 107 with LBP). The four groups (NP/LBP stratified for GP/PT) are largely comparable in gender, age,
161
fear of movement, pain catastrophizing and pain intensity, and the percentages of SBT-risk profiles of each of
162
the four groups are also comparable: the low-risk profile percentage ranges from 50.0% to 60.8%, the
ACCEPTED MANUSCRIPT 163
medium-risk profile percentage ranges from 35.1% to 42.3% and the high-risk profile percentage ranges from
164
4.1% to 11.2%. General practitioners tend to see a higher proportion of chronic patients with LBP compared to
165
PTs (68.8% vs. 48.6%).
166 Follow-up
168
The follow-up questionnaire for GPs and PTs had a non-response of 3.9% (n = 11), and the follow-up
169
questionnaire for patients had a failure to follow-up of 14.4% (n = 41). Thirteen patients were seen by both a
170
GP and a PT; in this analysis, these patients are analyzed in both groups. Figure 1 (patient flow) displays all
171
patients analyzed despite patients’ failure to follow-up.
172
Due to the clinicians’ non-response, the specific-agreement analysis was performed on 273 patients (175 with
173
LBP and 98 with NP), and due to the patients’ non-response, the persisting disability analysis was performed
174
on 243 patients (150 with LBP and 93 with NP). Due to both the clinicians’ and patients’ non-response, the
175
treatment analysis was performed on 234 patients (142 with LBP and 92 with NP).
M AN U
SC
RI PT
167
176 Treatments
178
The two authors independently categorized the treatments that the PTs applied, and, in 90.5% of the cases,
179
agreed on their categorization. The categorization of the GPs’ treatments was performed in SPSS. The authors
180
decided that the treatment offered by a GP, whereby he performed manipulations himself, is categorized as
181
medium-risk (comparable to referring the patient to a physiotherapist, irrespective of the number of
182
manipulations).
EP
AC C
183
TE D
177
184
Lower-back pain
185
In the specific-agreement analysis, the majority of the patients with LBP were at low-risk for persisting
186
disability (n = 95; 54.3%). Thirteen (13.7%) of these patients received the SBT-advised low-risk treatment
187
approach, which resulted in a specific agreement of 22.4% (Table 3). According to the SBT, all other patients (n
188
= 82) were over-treated. Receiving the treatment approach for medium or high risk did not result in a lower
189
percentage of persisting disability (Table 4a).
190
We found that of the 64 (36.8%) patients who were considered to be at medium risk, 55 (85.9%) received
191
treatment corresponding to this risk profile (specific agreement = 51.4%), seven patients received a low-risk
ACCEPTED MANUSCRIPT treatment (undertreated) and two a high-risk treatment (over-treated). Both over- and undertreated patients
193
had a higher percentage of persisting disability compared to the group that was treated in accordance with its
194
risk profile.
195
In the group of patients that were considered to be at high risk for persisting disability (n = 16; 9.2%), only one
196
patient (6.3%) received the high-risk treatment approach (specific agreement = 10.0%). According to the SBT,
197
all other patients (n = 15) were undertreated. We found that these undertreated patients had an average of
198
6.5 PT sessions, and the patients treated as high risk received 8.5 PT sessions on average.
RI PT
192
199 Neck pain
201
In the specific-agreement analysis, 57 (58.2%) of the 98 patients with NP were considered to be at low-risk for
202
persisting disability. Two of these patients (3.5%) received the corresponding low-risk treatment approach
203
(specific agreement = 6.3%); other patients were considered to be over-treated (Table 3) and had a higher
204
percentage of persisting disability at 3 months (Table 4b).
205
We found that while 36 patients (36.7%) were considered to be at medium risk for persisting disability, only 30
206
patients (83.3%) received care that corresponded to their medium-risk profile (specific agreement = 48.0%);
207
the remaining six patients were equally divided between receiving the low- and high-risk treatment approach.
208
Of the 98 patients with NP, the smallest group contains the patients considered to be at high risk for persisting
209
disability (N = 5; 5.1%). None of these patients received the high-risk treatment approach, making it impossible
210
to calculate the specific agreement. On average, these patients underwent 7.2 PT sessions, while patients
211
treated as high risk in this study received 8.8 sessions on average.
M AN U
TE D
EP
AC C
212
SC
200
213
Physiotherapy treatments
214
In the specific-agreement analysis, we found that 86.2% of the patients with LBP and 90.8% of the patients
215
with NP received a medium-risk treatment approach, which suggests a one-size-fits-all approach in current
216
usual care. Only seven patients (3.0%) received a high-risk approach, over half of the patients were over-
217
treated (47.9% for LBP and 59.8% for NP), and between 32.6% (NP) and 38.7% (LBP) received the targeted
218
treatment. The LBP and NP patients that were given the correct treatment had almost the same percentage of
219
persisting disability in 3 months (32.7% for LBP and 34.5% for NP).
220
ACCEPTED MANUSCRIPT General practitioners referral
222
The GPs referred 10 patients (9.9%) for imaging: seven to rule out serious pathology and three at the request
223
of the patient. Three other patients were referred to a neurologist or an orthopedic surgeon due to persisting
224
complaints; one of these referrals was at the request of the patient. In 27 cases, the GPs prescribed analgesic
225
drugs; for 15 patients (55.5%) this was time-contingent (compared to pain-contingent). Furthermore, we found
226
an (almost) absence of a psychosocial approach or related referrals by the GP (for example, the high-risk
227
approach).
RI PT
221
228 Discussion
230
Main findings
231
GPs and PTs treat the majority of their patients as medium-risk patients even though this majority is actually at
232
low risk for persisting complaints. No large differences are found between NP or LBP patients. The minority of
233
patients (36.7%) received care that corresponds to the advised care according to the SBT approach; 52.2% of
234
the patients were considered over-treated and 11.0% were undertreated.
M AN U
SC
229
235 Interpretation
237
The SBT approach advises one consultation with information or education for low-risk patients. We considered
238
an intervention to match the low-risk profile if the GP or PT gave information or education in either one or two
239
consultations. The GP also had the option to offer additional analgesics to the patient within the low-risk
240
treatment approach. The UK healthcare system advised an average of four interventions. In our cohort (LBP
241
and NP combined), we found an average of 6.8 PT sessions and 8.7 sessions in the patients with a high-risk
242
profile. Furthermore, we found an (almost) absence of a psychosocial approach or related referrals (for
243
example, the high-risk approach). This study clearly demonstrates that clinicians tend to treat the majority of
244
their patients as medium-risk patients; they do not stratify and treat patients based on the perceived risk of
245
persistent complaints, and over half of the patients are over-treated. In low-risk patients with LBP, we see that
246
the majority of the patients recover, irrespective of a low or medium-risk intervention. We expect the same for
247
NP; however, due to the small proportion of patients receiving a low-risk approach, we could not evaluate this.
248
Also, no conclusions can be drawn from the high-risk population due to the small sample size.
249
AC C
EP
TE D
236
ACCEPTED MANUSCRIPT 250
One Irish study conducted a non-randomized clinical trial in 332 LBP patients who were included in a historical
251
cohort. This group received a generic 12-week group-education or exercise program, comparable to the
252
medium-risk approach, while an intervention group of 251 LBP patients received a stratified-care approach.
253
Stratified care demonstrated a superior effect on the high-risk group; however, the study did not analyze the
254
usual care, as was done in this study.
255
In the original SBT study, stratified care was the intervention that was compared to a usual-care approach and
256
found that the intervention group displayed a higher mean change in disability compared to the control group;
257
however, no information about the specific contents of usual care was provided. No other studies have been
258
found that analyze un-protocolled usual care in contrast to the SBT approach.
259
Other remarkable findings are that the GPs referred 10 patients for imaging even though they were labeled as
260
non-specific LBP or NP, meaning that there were no signs or signals indicating serious pathology. The GP
261
guidelines advise against diagnostic imaging in these instances
262
healthcare costs due to additional tests and treatment, but may not improve clinical outcomes and may even
263
lower the quality of life.
264
basis, the GP guidelines advise doing so on a time-contingent basis.
RI PT
23
265
33,34
, since imagery may lead to higher
While analgesics were prescribed to 44.5% of the patients on a pain-contingent 33
TE D
35,36
M AN U
SC
22
Strengths and limitations
267
This is the first study that compares current usual care with the advised stratified-SBT approach in the
268
Netherlands, and provides insight into the question regarding whether specific implementation of stratified
269
care is required or whether usual care already properly categorizes patients.
270
Our results have limited generalizability for the group of high-risk patients because this group was quite small.
271
At the moment, the SBT approach is only available for LBP patients, and so far, no study has been conducted
272
on an SBT approach for patients with NP; however, the conditions are rather comparable. Our categorization
273
of usual care was based on the information that PTs provided. They wrote their (primary) treatment goals and
274
means, and we may have missed certain psychosocial factors that were addressed but not reported in the
275
patient files.
AC C
EP
266
276 277
In the UK, the SBT has the potential to predict persisting disability in clinical practice. We found differences
278
between usual care and the advised stratified care, since most patients received the therapy advised for
ACCEPTED MANUSCRIPT 279
medium-risk patients. Educating clinicians on the SBT approach, especially for the low and high-risk groups, is
280
necessary for this approach to be effective in reducing pain, function and sick leave in the Netherlands.
281
However, prior to making efforts to change the working method, it is essential to determine what causes
282
clinicians to treat patients as they do. A qualitative study should be undertaken to gain insight.
RI PT
283 Conclusion
285
Current Dutch usual care does not correspond to the advised stratified-care approach based on the SBT for
286
patients with non-specific LBP or NP. Clinicians tend to treat the majority of their patients as medium-risk
287
patients even though most of them were found to be low-risk patients.
SC
284
289
Acknowledgments
290
DELETED FOR REVIEWERS
M AN U
288
292
Authors’ contributions
293
DELETED FOR REVIEWERS
294 Conflict of Interest
296
DELETED FOR REVIEWERS
298 299
AC C
297
EP
295
TE D
291
ACCEPTED MANUSCRIPT 300
References
301
1.
Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289
302
diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010.
303
Lancet. 2012;380(9859):2163-2196. doi:10.1016/S0140-6736(12)61729-2. 2.
Côté P, Cassidy JD, Carroll L. The treatment of neck and low back pain: who seeks care? who goes
RI PT
304 305
where? Med Care. 2001;39(9):956-967. http://www.ncbi.nlm.nih.gov/pubmed/11502953. Accessed
306
October 22, 2015.
308 309
seeking. J Manipulative Physiol Ther. 2004;27(5):327-335. doi:10.1016/j.jmpt.2004.04.006. 4.
310 311
Walker BF, Muller R, Grant WD. Low back pain in Australian adults. Health provider utilization and care
SC
3.
Mannion AF, Wieser S, Elfering A. Association Between Beliefs and Care-Seeking Behavior for Low Back
M AN U
307
Pain. Spine (Phila Pa 1976). 2013;38(12):1016-1025. doi:10.1097/BRS.0b013e31828473b5. 5.
Kamper SJ, Maher CG, Hancock MJ, Koes BW, Croft PR, Hay E. Treatment-based subgroups of low back pain: A guide to appraisal of research studies and a summary of current evidence. Best Pract Res Clin
313
Rheumatol. 2010;24(2):181-191. doi:10.1016/j.berh.2009.11.003.
314
6.
TE D
312
van Tulder M, Becker A, Bekkering T, et al. Chapter 3 European guidelines for the management of acute nonspecific low back pain in primary care. Eur Spine J. 2006;15(S2):s169-s191.
316
doi:10.1007/s00586-006-1071-2. 7.
318 319
Bier JD, Scholten-Peeters W, Pool JJ, Staal JB, Tulder MW van, Verhagen AP. KNGF-Richtlijn Nekpijn.; 2016.
8.
320
AC C
317
EP
315
Koes BW, van Tulder M, Lin C-WC, Macedo LG, McAuley J, Maher C. An updated overview of clinical guidelines for the management of non-specific low back pain in primary care. Eur Spine J.
321
2010;19(12):2075-2094. doi:10.1007/s00586-010-1502-y.
322
9.
NHG. NHG-Standaard Aspecifieke Lagerugpijn.; 2005.
323
10.
Staal JB, Hendriks EJM, Heijmans M, et al. KNGF-richtlijn Lage rugpijn. 2013.
324
11.
Hill JC, Dunn KM, Lewis M, et al. A primary care back pain screening tool: Identifying patient subgroups
ACCEPTED MANUSCRIPT 325 326
for initial treatment. Arthritis Rheum. 2008;59(5):632-641. doi:10.1002/art.23563. 12.
327 328
Foster NE, Hill JC, Hay EM. Subgrouping patients with low back pain in primary care: are we getting any better at it? Man Ther. 2011;16(1):3-8. doi:10.1016/j.math.2010.05.013.
13.
Abedi M, Manshadi FD, Khalkhali M, et al. Translation and validation of the Persian version of the STarT Back Screening Tool in patients with nonspecific low back pain. Man Ther. 2015:1-5.
330
doi:10.1016/j.math.2015.04.006. 14.
332 333
Pilz B, Vasconcelos RA, Marcondes FB. The Brazilian version of STarT Back Screening Tool – translation , cross-cultural adaptation and reliability *. 2014;18(5):453-461.
15.
SC
331
RI PT
329
Luan S, Min Y, Li G, et al. Cross-cultural Adaptation, Reliability, and Validity of the Chinese Version of the STarT Back Screening Tool in Patients With Low Back Pain. Spine (Phila Pa 1976). 2014;39(16):E974-
335
9. doi:10.1097/BRS.0000000000000413.
336
16.
M AN U
334
Piironen S, Paananen M, Haapea M, et al. Transcultural adaption and psychometric properties of the STarT Back Screening Tool among Finnish low back pain patients. Eur Spine J. February 2015.
338
doi:10.1007/s00586-015-3804-6.
339
17.
TE D
337
Bruyere O, Demoulin M, Brereton C, et al. Translation validation of a new back pain screening questionnaire (the STarT Back Screening Tool) in French. Arch Public Heal. 2012;70(1):12.
341
doi:10.1186/0778-7367-70-12. 18.
343
Back Screening Tool” (SBST) in different subgroups. Aten primaria / Soc Española Med Fam y
344 345
Comunitaria. 2011;43(7):356-361. doi:10.1016/j.aprim.2010.05.019.
19.
346 347
350
Morso L, Albert H, Kent P, et al. Translation and discriminative validation of the STarT Back Screening Tool into Danish. Eur Spine J. 2011;20(12):2166-2173. doi:10.1007/s00586-011-1911-6.
20.
348 349
Gusi N, del Pozo-Cruz B, Olivares PR, Hernández-Mocholi M, Hill JC. The Spanish version of the “STarT
AC C
342
EP
340
Bier JD, Ostelo RWJG, Hooff ML van, et al. Validity and reproducibility of the STarT Back Tool (Dutch version) in low back pain patients in primary care. PTJ - pending Publ. 2016.
21.
Bier JD, Ostelo RWJG, Koes BW, Verhagen AP. Validity and reproducibility of the modified STarT Back Tool (Dutch version) for patients with neck pain in primary care. Unpubl Obs. 2016.
ACCEPTED MANUSCRIPT 351
22.
Hill JC, Whitehurst DGT, Lewis M, et al. Comparison of stratified primary care management for low
352
back pain with current best practice (STarT Back): a randomised controlled trial. Lancet (London,
353
England). 2011;378(9802):1560-1571. doi:10.1016/S0140-6736(11)60937-9.
354
23.
Murphy SE, Blake C, Power CK, Fullen BM. Comparison of a Stratified Group Intervention (STarT Back) With Usual Group Care in Patients With Low Back Pain: A Nonrandomized Controlled Trial. Spine (Phila
356
Pa 1976). 2016;41(8):645-652. doi:10.1097/BRS.0000000000001305.
357
24.
RI PT
355
Hjermstad MJ, Fayers PM, Haugen DF, et al. Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature
359
review. J Pain Symptom Manag. 2011;41(6):1073-1093. doi:10.1016/j.jpainsymman.2010.08.016. 25.
361 362
Vernon H, Mior S. The Neck Disability Index: a study of reliability and validity. J Manip Physiol Ther.
M AN U
360
SC
358
1991;14(7):409-415. http://www.ncbi.nlm.nih.gov/pubmed/1834753. 26.
Brouwer S, Kuijer W, Dijkstra PU, Goeken LN, Groothoff JW, Geertzen JH. Reliability and stability of the Roland Morris Disability Questionnaire: intra class correlation and limits of agreement. Disabil Rehabil.
364
2004;26(3):162-5. doi:10.1080/09638280310001639713.
365
27.
TE D
363
Roland M, Morris R. A study of the natural history of low-back pain. Part II: development of guidelines for trials of treatment in primary care. Spine (Phila Pa 1976). 1983;8(2):145-150. doi:10.1007/978-1-
367
4471-5451-8_59. 28.
369
pain and its relation to behavioral performance. Pain. 1995;62(3):363-372.
370 371
http://www.ncbi.nlm.nih.gov/pubmed/8657437.
29.
372 373
Vlaeyen JW, Kole-Snijders AM, Boeren RG, van Eek H. Fear of movement/(re)injury in chronic low back
AC C
368
EP
366
Sullivan MJL, Bishop SR, Pivik J. The Pain Catastrophizing Scale: Development and validation. Psychol
Assess. 1995;7(4):524-532. doi:Doi 10.1037/1040-3590.7.4.524.
30.
Salen BA, Spangfort E V, Nygren AL, Nordemar R. The Disability Rating Index: an instrument for the
374
assessment of disability in clinical settings. J Clin Epidemiol. 1994;47(12):1423-1435.
375
http://www.ncbi.nlm.nih.gov/pubmed/7730851.
376
31.
Hay EM, Dunn KM, Hill JC, et al. A randomised clinical trial of subgrouping and targeted treatment for
ACCEPTED MANUSCRIPT 377
low back pain compared with best current care. The STarT Back Trial Study Protocol. BMC
378
Musculoskelet Disord. 2008;9:58. doi:10.1186/1471-2474-9-58. 32.
380 381
de Vet HCW, Mokkink LB, Terwee CB, Hoekstra OS, Knol DL. Clinicians are right not to like Cohen’s kappa. Bmj-British Med J. 2013;346(April):f2125. doi:Artn F2125Doi 10.1136/Bmj.F2125.
33.
Chavannes AW, Mens JMA, Koes BW, et al. NHG-Standaard Aspecifieke lagerugpijn | NHG. Huisarts
RI PT
379
382
Wet. 2005;48(3):113-123. https://www.nhg.org/standaarden/volledig/nhg-standaard-aspecifieke-
383
lagerugpijn#note-15. Accessed April 22, 2016. 34.
van Tulder MW, Koes BW. Evidence-Based Handelen Bij Lage Rugpijn.; 2013.
SC
384
http://download.springer.com/static/pdf/687/bfm%25253A978-90-368-0277-
386
2%25252F1.pdf?originUrl=http%253A%252F%252Flink.springer.com%252Fbook%252Fbfm%253A978-
387
90-368-0277-
388
2%252F1&token2=exp=1461316561~acl=%252Fstatic%252Fpdf%252F687%252Fbfm%2525253A978-
389
90-368-0277-2%2525252F1.pdf%253Forigi. Accessed April 22, 2016.
390
35.
M AN U
385
Jarvik JG, Hollingworth W, Martin B, et al. Rapid magnetic resonance imaging vs radiographs for patients with low back pain: a randomized controlled trial. JAMA. 2003;289(21):2810-2818.
392
doi:10.1001/jama.289.21.2810.
393
36.
TE D
391
Karel YHJM, Verkerk K, Endenburg S, Metselaar S, Verhagen AP. Effect of routine diagnostic imaging for patients with musculoskeletal disorders: A meta-analysis. Eur J Intern Med. 2015;26(8):585-595.
395
doi:10.1016/j.ejim.2015.06.018.
397
AC C
396
EP
394
ACCEPTED MANUSCRIPT Figure 1; patient flow
399
Table 1; specific agreement
400
Table 2; baseline characteristics
401
Table 3; specific agreement analysis between baseline risk profile and treatment as provided
402
Table 4a; treatment analysis for low back pain
403
Table 4b; treatment analysis for neck pain
RI PT
398
AC C
EP
TE D
M AN U
SC
404
ACCEPTED MANUSCRIPT Table 1, specific agreement*
Low
Medium
High
Low
(A)
(B)
(C)
Medium
(D)
(E)
(F)
High
(G)
(H)
(I)
RI PT
* “Low-risk” A/(A+(B+C+D+G)/2) “Medium-risk” E/(E+(B+H+D+F)/2)
EP
TE D
M AN U
SC
“High-risk” I/(I+(C+F+G+H)/2)
AC C
Baseline (T0)
Follow-up (T1)
ACCEPTED MANUSCRIPT Table 2, baseline characteristics* Total
GP
Neck (n=100) Back (n=184)
†
PT
Neck (n=26)
Back (n=77)
‡
Neck (n=74) Back (n=107)
Female
65 (65.0)
103 (56.0)
17 (65.4)
42 (54.5)
48 (64.9)
61 (57.0)
Age in years, mean (SD)
45.6 (14.3)
44.7 (14.6)
45.4 (12.5)
40.7 (14.7)
45.7 (14.9)
47.5 (13.9)
Low
58 (58.0)
96 (52.2)
13 (50.0)
42 (54.5)
45 (60.8)
54 (50.5)
Medium
37 (37.0)
70 (38.0)
11 (42.3)
29 (37.7)
26 (35.1)
41 (38.3)
High
5 (5.0)
18 (9.8)
2 (7.7)
6 (7.8)
3 (4.1)
12 (11.2)
<1 month
27 (27.0)
53 (28.8)
7 (26.9)
15 (19.5)
20 (27.0)
38 (35.5)
1 to 3 months
18 (18.0)
26 (14.1)
5 (19.2)
9 (11.7)
13 (17.6)
17 (15.9)
>3 months
55 (55.0)
105 (57.1)
14 (53.8)
53 (68.8)
41 (55.4)
52 (48.6)
||
5.5 (1.9)
5.9 (1.8)
5.4 (1.9)
5.7 (1.9)
§
¶
Disability (NDI) , mean (SD)
14.1 (6.8)
#
9.5 (5.9)
Disability (RDQ) , mean (SD) **
Fear (TSK) , mean (SD)
14.2 (8.8)
32.2 (5.9)
††
Catastrophizing (PCS) , mean (SD) 12.7 (10.0)
6.0 (1.8)
14.1 (6.1)
8.9 (5.7)
10.0 (6.1)
34.8 (7.1)
33.8 (5.6)
35.2 (6.3)
31.7 (5.9)
34.4 (7.6)
13.8 (10.3)
16.5 (12.4)
14.6 (11.2)
11.4 (8.7)
13.1 (9.6)
17 (23.6)
23 (23.2)
EP
TE D
Referral * Values are numbers (percentage) unless otherwise indicated. † General practitioner ‡ Physical therapist § SBT = STarT Back Tool (0-9) || Pain intensity is measured on a Numeric Pain Rating Scale (0-10) ¶ NDI = Neck Disability Index (0-50) # RDQ = Roland Disability Questionnaire (0-24) ** TSK = Tampa Scale of Kinesiophobia (17-63) †† PCS = Pain Catastrophizing Scale (0-65)
AC C
5.6 (1.9)
M AN U
Pain intensity , mean (SD)
SC
Episode duration
RI PT
SBT risk profile
ACCEPTED MANUSCRIPT Table 3, Specific agreement analysis between baseline risk profile and treatment as provided*
Treatment profile Medium
High
13 7 1
81 55 14
1 2 1
22.4% 51.4% 10.0%
2 3 1
55 30 4
0 3 0
6.3% 48.0% N/A
RI PT
Neck pain (n=98) Low-risk Medium-risk High-risk
Specific Agreement
Low
SC
Baseline Low Back pain (n=175) Low-risk Medium-risk High-risk
AC C
EP
TE D
M AN U
* Analysis based on patients’ baseline-data and clinicians treatment-data. Specific agreement is interpreted as the proportion of patients that had that specific risk profile on either baseline or in the treatment profile.
ACCEPTED MANUSCRIPT
Mediumrisk High-risk
Under treated Rightfully treated Over treated
N (%) 9 (12.0) 65 (86.7) 1 (1.3) 6 (11.3) 45 (84.9) 2 (3.8) 0 (0.0) 13 (92.9) 1 (7.1) 19 (13.4) 55 (38.7) 68 (47.9)
SC
Treatment Low Medium High Low Medium High Low Medium High
M AN U
Low Back pain N=142
Baseline Low-risk
Persisting Disability N (%) 2 (22.2) 13 (20.0) 1 (100.0) 3 (50.0) 15 (33.3) 2 (100.0) 0 (N/A) 7 (53.8) 1 (100.0) 10 (52.6) 18 (32.7) 16 (23.5)
RI PT
Table 4a, treatment analysis for low back pain*
* Analysis based on patients’ baseline-data and follow-up data, and clinicians treatment-data. ‘Persisting Disability’ is the amount and percentage of patients with persisting disability in the corresponding treatment group
Neck pain N=92
Baseline Low-risk
Treatment Low Medium High Low Medium High Low Medium High
EP
Mediumrisk
TE D
Table 4b, treatment analysis for low neck pain*
AC C
High-risk
Under treated Rightfully treated Over treated
(%) 2 (3.7) 52 (96.3) 0 (0.0) 3 (8.8) 28 (82.4) 3 (8.8) 1 (25.0) 3 (75.0) 0 (0.0) 7 (7.6) 30 (32.6) 55 (59.8)
Persisting Disability N (%) 0 (0.0) 7 (13.5) 9 (N/A) 0(0.0) 10 (35.7) 3 (100.0) 1 (100.0) 2 (66.7) 0 (N/A) 3 (42.9) 10 (33.3) 19 (34.5)
* Analysis based on patients’ baseline-data and follow-up data, and clinicians treatment-data. ‘Persisting Disability’ is the amount and percentage of patients with persisting disability in the corresponding treatment group
ACCEPTED MANUSCRIPT Figure 1; Patient flow
Back pain N=184
Neck pain N=100
RI PT
Baseline data N=284
PT N=100
Analyzed PT’s N=108
GP N=21
Analyzed GP’s N=26
AC C
EP
TE D
Analyzed GP’s N=75
Both N=8
Both N=5
PT N=72
M AN U
GP N=67
SC
No GP/PT Follow-up N=2
No GP/PT Follow-up N=9
Analyzed PT’s N=77
S0003-9993(17)30456-2
DOI:
10.1016/j.apmr.2017.06.011
Reference:
YAPMR 56947
To appear in:
ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION
Received Date: 14 November 2016 Revised Date:
31 May 2017
Accepted Date: 5 June 2017
Please cite this article as: Bier JD, Sandee-Geurts JJW, Ostelo RWJG, Koes BW, Verhagen AP, Can primary care for back and/or neck pain in the Netherlands benefit from stratification for risk groups according to the STarT Back Tool-classification?, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2017), doi: 10.1016/j.apmr.2017.06.011. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Primary care for back and/or neck pain
ACCEPTED MANUSCRIPT
Can primary care for back and/or neck pain in the Netherlands benefit from stratification for risk groups according to the STarT Back Tool-classification?
RI PT
Authors: 1, 2
Jasper D. Bier, MSc
Janneke J.W. Sandee-Geurts, MSc
Bart W. Koes, Prof
5,6
SC
Raymond W. J. G. Ostelo, Prof
3,4
1
1
M AN U
Arianne P. Verhagen, PhD
Department of General Practice, Erasmus MC, PO Box 2040, 3000 CA, Rotterdam, The Netherlands
2.
FS Fysio, Capelle aan den IJssel, The Netherlands
3.
Faculty of Human Movement Sciences, VU, Amsterdam, The Netherlands
4.
Fysiotherapie Kapellaan / Ouwerkerk, Vught, The Netherlands
5.
Department of Epidemiology and Biostatistics, VU University Medical Centre Amsterdam and the
TE D
1.
EMGO Institute for Health and Care Research, The Netherlands 6.
Department of Health Sciences, Faculty of Earth and Life Sciences, VU University Amsterdam, The
Correspondence to:
EP
Netherlands.
AC C
Jasper D. Bier, Department of General Practice, Erasmus MC, PO Box 2040, 3000CA Rotterdam, The Netherlands, T +31 6 26536787
E [email protected]
URL www.erasmusmc.nl
The study was approved by the medical ethics committee of the Erasmus University, Rotterdam, The Netherlands. (METC-2014-256)
The authors declare that there is no conflict of interests regarding the publication of this paper. This study was undertaken with financial support of CZ healthcare insurance company and Dutch Arthritis Association.
Primary care for back and/or neck pain
ACCEPTED MANUSCRIPT
The authors would like to thank all of the GP’s and PT’s that included patients for this study. A special thank goes out to Nynke Wildervanck for her contribution and to Steven Constandse, Guido Iken, Joost van
AC C
EP
TE D
M AN U
SC
RI PT
Broekhoven and Frans van der Kooij for their extra efforts to reach the needed sample size.
ACCEPTED MANUSCRIPT 1
Titel:
2
Can primary care for back and/or neck pain in the Netherlands benefit from stratification for
3
risk groups according to the STarT Back Tool-classification?
4 Abstract
6
Objective: To evaluate whether current Dutch primary-care clinicians offer tailored treatment to patients with
7
lower-back pain (LBP) or neck pain (NP) according to their risk stratification, based on the Keele STarT
8
(Subgroup Targeted Treatment) Back-Screening Tool (SBT).
9
Design: Prospective cohort study with 3 month follow-up
SC
RI PT
5
Setting: Primary care
11
Participants: General practitioners (GPs) and physiotherapists (PTs) included patients with non-specific LBP
12
and/or NP.
13
Interventions: Patients completed a baseline questionnaire, including the Dutch SBT, for either LBP or NP. A
14
follow-up measurement was conducted after 3 months to determine recovery (using the Global Perceived
15
Effect (GPE) scale), pain (using the Numeric Pain-Rating Scale (NPRS)) and function (using the Roland Disability
16
Questionnaire (RDQ) or the Neck Disability Index (NDI)). A questionnaire was sent to the GPs and PTs to
17
evaluate the provided treatment.
18
Main outcome measures: Prevalence of patients’ risk profile and clinicians’ applied care, and the percentage
19
of patients with persisting disability at follow-up. A distinction was made between patients receiving the
20
advised treatment and those receiving the non-advised treatment.
21
Results: In total, 12 GPs and 33 PTs included patients. After 3 months, we analyzed 184 patients with LBP and
22
100 patients with NP. In the LBP group, 52.2% of the patients were at low risk for persisting disability, 38.0%
23
were at medium risk and 9.8% were at high risk. Overall, 24.5% of the LBP patients received a low-risk
24
treatment approach, 73.5% a medium-risk and 2.0% a high-risk treatment approach. The specific agreement
25
between the risk profile and the received treatment for patients with LBP was poor for the low-risk and high-
26
risk patients (respectively 21.1% and 10.0%), and fair for medium-risk patients (51.4%). In the NP group, 58.0%
27
of the patients were at low risk for persisting disability, 37.0% were at medium risk and 5.0% were at high-risk.
28
Only 6.1% of the patients with NP received the low-risk treatment approach. The medium-risk treatment
AC C
EP
TE D
M AN U
10
ACCEPTED MANUSCRIPT approach was offered the most (90.8%) and the high-risk approach was applied in only 3.1% of the patients.
30
The specific agreement between the risk profile and received treatment for NP patients was poor for low-risk
31
and medium-risk patients (resp. 6.3% and 48.0%); agreement for high-risk patients could not be calculated.
32
Conclusion: Current Dutch primary care for patients with non-specific LBP and/or NP does not correspond to
33
the advised stratified-care approach based on the SBT as the majority of patients receive medium risk
34
treatment. The majority of “low-risk” patients are over-treated and the majority of “high-risk” patients are
35
undertreated. Although the stratified-care approach has not yet been validated in Dutch primary care, these
36
results indicate that there may be substantial room for improvement.
37
Keywords: Usual care, general practitioner, physical therapy, lower-back pain, neck pain, STarT Back Tool
38
Abbreviations:
39
General Practitioner (GP)
40
Global Perceived Effect (GPE)
41
Lower-back pain (LBP)
42
Neck Disability Index (NDI)
43
Neck Pain (NP)
44
Numeric Pain-Rating Scale (NPRS)
45
Roland Disability Questionnaire (RDQ)
46
STarT Back-Screening Tool (SBT)
47
Subgroup Targeted Treatment (STarT)
AC C
EP
TE D
M AN U
SC
RI PT
29
ACCEPTED MANUSCRIPT 48
Introduction
49
Lower-back pain and NP are major public-health problems; they are the primary and fourth causes,
50
respectively, of disability worldwide. With regard to health-seeking behavior for back and neck pain,
51
approximately 55% of patients seek healthcare, and between 12% and 32% visit a GP.
52
Non-specific LBP and NP are the main focus of most primary-care guidelines.
53
Dutch General-Practitioners Society and the Royal Dutch Physiotherapists Society divide patients with non-
54
specific LBP and NP into roughly two subgroups: 1) normal recovery, which is defined as a decrease in pain and
55
function limitation prior to consultation, and 2) a (suspected) delayed recovery.
56
The SBT is a tool, developed in England, to allocate primary-care patients with LBP to three prognostic
57
subgroups: patients at either low, medium or high risk for persisting disability. This allocation aims to ensure
58
that the appropriate stratified care is applied to patients “at risk” for persistent LBP in order to prevent it.
59
This tool’s questions are based on known negative-prognostic factors that can be influenced by treatment.
60
Furthermore, the SBT has been found to be a valid and reliable tool for subgrouping patients in the UK , and it
61
has been translated and validated in several languages, including Dutch , since its initial English publication in
62
2008.
63
corresponding targeted treatment, has yet to be implemented in the Netherlands.
64
Advised care for normal recovery is to reassure the patient and inform him or her of the positive prognosis,
65
and to advise the patient to stay active, which might be supported with the prescription of pain medication. If
66
the pain persists or worsens, the GP can refer physiotherapy to the patient. Where current guidelines advise a
67
stepped approach for patients with a delayed recovery, the SBT advises a stratified-care approach; the
68
difference between these two approaches relates to the timing of (effective) interventions. Implementing the
69
SBT in the UK has led to higher quality-adjusted life years for patients and lower health costs, and superior
70
outcomes in the high-risk group using stratified care.
71
patients with NP; however, we expect that it will not differ from the approach for LBP. Before advising on
72
implementing the SBT in the Netherlands, we need more insight into the current usual care because
73
implementation of the SBT might be unnecessary if usual care provides comparable or better outcomes.
74
Therefore, the aim of this study is to evaluate current Dutch primary care for spinal pain and whether it
75
corresponds to the advised stratified-care approach based on the SBT.
76
1
The guidelines from both the
RI PT
7,8
2–4
SC
7,9,10
M AN U
11
12
11
20
21
The Dutch version of the SBT has been modified to fit patients with NP. The SBT, and
AC C
EP
TE D
13–19
22,23
No studies have been published on stratified care for
ACCEPTED MANUSCRIPT Method
78
Design
79
The prevalence of risk groups in neck- and back-pain patients according to the SBT (PRINS) study is a
80
prospective cohort study, which includes patients with LBP or NP of any duration that consulted a GP or PT.
81
The study was approved by the medical ethics committee of the Erasmus University, Rotterdam, the
82
Netherlands (METC-2014-256).
RI PT
77
83 Participants
85
Care providers
86
We asked GPs and PTs, who work in the primary-care sector and had displayed interest in the SBT during pilot
87
projects, to participate in the PRINS study. Information about the study protocol was provided through several
88
meetings, by phone or by digital/paper documentation, and participating GPs and PTs received the study
89
protocol and a folder with patient information and informed consent forms. The majority of these GPs and PTs
90
work in small clinics in the region of Rotterdam (a maximum of 50 kilometers), the Netherlands.
M AN U
SC
84
91 Patients
93
The inclusion period for patients started in November 2014 and continued through to May 2015, and patients
94
consulting their PT or GP for LBP or NP during that period were invited to participate. Other inclusion criteria
95
were that the patient had to be 18 years old or over, could speak and read Dutch and had an email address.
96
Patients were excluded if, during the consultation, the GP or PT found red flags indicating a possible serious
97
underlying pathology (for example, an infection, a fracture, cauda equina or a tumor) responsible for the LBP
98
or NP.
99
Patients were provided with oral and written information about the aim of the study and the procedure of
100
data collection, and each patient signed an informed consent, which was handed back to the PT or GP who
101
subsequently registered the patient online. The patient immediately received an email with a link to the
102
baseline questionnaire, and if necessary, a reminder to complete the questionnaire was sent after a few days.
103
Patients who did not complete the baseline questionnaire within 7 days were excluded from the cohort.
EP
AC C
104 105
TE D
92
Treatment
ACCEPTED MANUSCRIPT 106
The patients received usual care from their GPs or PTs, who were kept blind for the results of the patients’
107
baseline questionnaire.
108 Baseline measurements
110
The baseline questionnaire consisted of questions on demographic data, the SBT (either the back or neck
111
version) and the NPRS to assess pain; the NDI or the RDQ
112
kinesiophobia; and the pain-catastrophizing scale and the EQ-5d to assess quality of life. The NPRS options
113
range from 0 (no pain) to 10 (worst imaginable pain), the RDQ consists of 24 statements with a “yes” or “no”
114
answer option and a total score ranging from 0 to 24, and the NDI consists of 10 statements with a 6-point
115
scale ranging from 0 (not limited) to 6 (completely limited) and a total score ranging from 0 to 50.
24
25
28
to assess disability; the Tampa scale for 30
SC
29
26,27
RI PT
109
M AN U
116 Outcome measure
118
Three months after inclusion, the patients received a follow-up questionnaire to assess pain (NPRS), disability
119
(RDQ or NDI) and recovery, using the GPE. The answer options on the GPE range from 1 (fully recovered) to 7
120
(worse than ever).
121
Persisting LBP disability was defined as an RDQ score ≥ 7, based on the mean of the baseline score as used by
122
Hill at al. , while persisting NP disability was defined as an NDI score ≥ 13, based on the median baseline score
123
in this cohort.
124
Three months after inclusion, the GP was sent a questionnaire to inquire about the number of visits,
125
prescribed medication, referrals to PTs or medical professionals and requested imaging and blood tests. The PT
126
received a questionnaire to inquire about treatment data such as the number and period of treatments, and
127
the aim and means of treatment. The investigator sent and received all questionnaires digitally, and they were
128
handled and stored through LimeSurvey 2.05.
EP
11
AC C
129
TE D
117
130
Analysis
131
Treatment categorization
132
The two authors utilized the following criteria to independently categorize the treatments applied by the GPs
133
and/or PTs (in case of differences between the authors, the categories were determined by consensus):
ACCEPTED MANUSCRIPT 134
•
Low-risk approach – the GP provided information, advice and some analgesics or one or two PT
135
consultations, and the treatment was hands-off and consisted of offering information, advice and
136
exercises. •
Medium-risk approach – in addition to the low-risk approach, the GP referred the patient to a PT, and
138 139
the PT performed an evidence-based intervention. •
RI PT
137
High-risk approach – in addition to the medium-risk approach, the GP referred the patient to either a
140
PT specialized in treating patients with a psychosomatic approach, a psychologist or equivalent, and
141
the PT assessed bio-psychosocial risk factors and used cognitive behavioral principles as
142
interventions.
SC
31
143 Statistical analysis
145
We described the characteristics of the clinicians and patient population using frequencies (means with
146
standard deviations). Next, we calculated the frequencies of patients per SBT-risk profile and the clinicians’
147
treatment approaches. The specific agreement calculates percentages of advised-treatment approaches for
148
each risk profile separately. For example, patients who were “low-risk” at baseline and treated as such are
149
calculated as a proportion of patients that were “low-risk” on either of the two measurements. We modified
150
the specific agreement to fit a 3x3 table, as illustrated in Table 1, because the original method is done in a 2x2
151
table. We rated a specific agreement < 40% as poor, 40% to 59% as fair, 60% to 74% as good and 75% to 100%
152
as excellent. Lastly, we calculated the percentage of patients with persisting disability per risk profile; this was
153
done for the groups receiving the advised and non-advised treatments.
EP
TE D
32
AC C
154
M AN U
144
155
Results
156
Study population
157
The GPs and PTs originally included 370 patients; however, 86 patients failed to fill in the baseline
158
questionnaire and were excluded. The characteristics of the 284 remaining patients are presented in Table 2.
159
The 12 GPs included 103 patients (26 with NP and 77 with LBP) and the 33 PTs included 181 patients (74 with
160
NP and 107 with LBP). The four groups (NP/LBP stratified for GP/PT) are largely comparable in gender, age,
161
fear of movement, pain catastrophizing and pain intensity, and the percentages of SBT-risk profiles of each of
162
the four groups are also comparable: the low-risk profile percentage ranges from 50.0% to 60.8%, the
ACCEPTED MANUSCRIPT 163
medium-risk profile percentage ranges from 35.1% to 42.3% and the high-risk profile percentage ranges from
164
4.1% to 11.2%. General practitioners tend to see a higher proportion of chronic patients with LBP compared to
165
PTs (68.8% vs. 48.6%).
166 Follow-up
168
The follow-up questionnaire for GPs and PTs had a non-response of 3.9% (n = 11), and the follow-up
169
questionnaire for patients had a failure to follow-up of 14.4% (n = 41). Thirteen patients were seen by both a
170
GP and a PT; in this analysis, these patients are analyzed in both groups. Figure 1 (patient flow) displays all
171
patients analyzed despite patients’ failure to follow-up.
172
Due to the clinicians’ non-response, the specific-agreement analysis was performed on 273 patients (175 with
173
LBP and 98 with NP), and due to the patients’ non-response, the persisting disability analysis was performed
174
on 243 patients (150 with LBP and 93 with NP). Due to both the clinicians’ and patients’ non-response, the
175
treatment analysis was performed on 234 patients (142 with LBP and 92 with NP).
M AN U
SC
RI PT
167
176 Treatments
178
The two authors independently categorized the treatments that the PTs applied, and, in 90.5% of the cases,
179
agreed on their categorization. The categorization of the GPs’ treatments was performed in SPSS. The authors
180
decided that the treatment offered by a GP, whereby he performed manipulations himself, is categorized as
181
medium-risk (comparable to referring the patient to a physiotherapist, irrespective of the number of
182
manipulations).
EP
AC C
183
TE D
177
184
Lower-back pain
185
In the specific-agreement analysis, the majority of the patients with LBP were at low-risk for persisting
186
disability (n = 95; 54.3%). Thirteen (13.7%) of these patients received the SBT-advised low-risk treatment
187
approach, which resulted in a specific agreement of 22.4% (Table 3). According to the SBT, all other patients (n
188
= 82) were over-treated. Receiving the treatment approach for medium or high risk did not result in a lower
189
percentage of persisting disability (Table 4a).
190
We found that of the 64 (36.8%) patients who were considered to be at medium risk, 55 (85.9%) received
191
treatment corresponding to this risk profile (specific agreement = 51.4%), seven patients received a low-risk
ACCEPTED MANUSCRIPT treatment (undertreated) and two a high-risk treatment (over-treated). Both over- and undertreated patients
193
had a higher percentage of persisting disability compared to the group that was treated in accordance with its
194
risk profile.
195
In the group of patients that were considered to be at high risk for persisting disability (n = 16; 9.2%), only one
196
patient (6.3%) received the high-risk treatment approach (specific agreement = 10.0%). According to the SBT,
197
all other patients (n = 15) were undertreated. We found that these undertreated patients had an average of
198
6.5 PT sessions, and the patients treated as high risk received 8.5 PT sessions on average.
RI PT
192
199 Neck pain
201
In the specific-agreement analysis, 57 (58.2%) of the 98 patients with NP were considered to be at low-risk for
202
persisting disability. Two of these patients (3.5%) received the corresponding low-risk treatment approach
203
(specific agreement = 6.3%); other patients were considered to be over-treated (Table 3) and had a higher
204
percentage of persisting disability at 3 months (Table 4b).
205
We found that while 36 patients (36.7%) were considered to be at medium risk for persisting disability, only 30
206
patients (83.3%) received care that corresponded to their medium-risk profile (specific agreement = 48.0%);
207
the remaining six patients were equally divided between receiving the low- and high-risk treatment approach.
208
Of the 98 patients with NP, the smallest group contains the patients considered to be at high risk for persisting
209
disability (N = 5; 5.1%). None of these patients received the high-risk treatment approach, making it impossible
210
to calculate the specific agreement. On average, these patients underwent 7.2 PT sessions, while patients
211
treated as high risk in this study received 8.8 sessions on average.
M AN U
TE D
EP
AC C
212
SC
200
213
Physiotherapy treatments
214
In the specific-agreement analysis, we found that 86.2% of the patients with LBP and 90.8% of the patients
215
with NP received a medium-risk treatment approach, which suggests a one-size-fits-all approach in current
216
usual care. Only seven patients (3.0%) received a high-risk approach, over half of the patients were over-
217
treated (47.9% for LBP and 59.8% for NP), and between 32.6% (NP) and 38.7% (LBP) received the targeted
218
treatment. The LBP and NP patients that were given the correct treatment had almost the same percentage of
219
persisting disability in 3 months (32.7% for LBP and 34.5% for NP).
220
ACCEPTED MANUSCRIPT General practitioners referral
222
The GPs referred 10 patients (9.9%) for imaging: seven to rule out serious pathology and three at the request
223
of the patient. Three other patients were referred to a neurologist or an orthopedic surgeon due to persisting
224
complaints; one of these referrals was at the request of the patient. In 27 cases, the GPs prescribed analgesic
225
drugs; for 15 patients (55.5%) this was time-contingent (compared to pain-contingent). Furthermore, we found
226
an (almost) absence of a psychosocial approach or related referrals by the GP (for example, the high-risk
227
approach).
RI PT
221
228 Discussion
230
Main findings
231
GPs and PTs treat the majority of their patients as medium-risk patients even though this majority is actually at
232
low risk for persisting complaints. No large differences are found between NP or LBP patients. The minority of
233
patients (36.7%) received care that corresponds to the advised care according to the SBT approach; 52.2% of
234
the patients were considered over-treated and 11.0% were undertreated.
M AN U
SC
229
235 Interpretation
237
The SBT approach advises one consultation with information or education for low-risk patients. We considered
238
an intervention to match the low-risk profile if the GP or PT gave information or education in either one or two
239
consultations. The GP also had the option to offer additional analgesics to the patient within the low-risk
240
treatment approach. The UK healthcare system advised an average of four interventions. In our cohort (LBP
241
and NP combined), we found an average of 6.8 PT sessions and 8.7 sessions in the patients with a high-risk
242
profile. Furthermore, we found an (almost) absence of a psychosocial approach or related referrals (for
243
example, the high-risk approach). This study clearly demonstrates that clinicians tend to treat the majority of
244
their patients as medium-risk patients; they do not stratify and treat patients based on the perceived risk of
245
persistent complaints, and over half of the patients are over-treated. In low-risk patients with LBP, we see that
246
the majority of the patients recover, irrespective of a low or medium-risk intervention. We expect the same for
247
NP; however, due to the small proportion of patients receiving a low-risk approach, we could not evaluate this.
248
Also, no conclusions can be drawn from the high-risk population due to the small sample size.
249
AC C
EP
TE D
236
ACCEPTED MANUSCRIPT 250
One Irish study conducted a non-randomized clinical trial in 332 LBP patients who were included in a historical
251
cohort. This group received a generic 12-week group-education or exercise program, comparable to the
252
medium-risk approach, while an intervention group of 251 LBP patients received a stratified-care approach.
253
Stratified care demonstrated a superior effect on the high-risk group; however, the study did not analyze the
254
usual care, as was done in this study.
255
In the original SBT study, stratified care was the intervention that was compared to a usual-care approach and
256
found that the intervention group displayed a higher mean change in disability compared to the control group;
257
however, no information about the specific contents of usual care was provided. No other studies have been
258
found that analyze un-protocolled usual care in contrast to the SBT approach.
259
Other remarkable findings are that the GPs referred 10 patients for imaging even though they were labeled as
260
non-specific LBP or NP, meaning that there were no signs or signals indicating serious pathology. The GP
261
guidelines advise against diagnostic imaging in these instances
262
healthcare costs due to additional tests and treatment, but may not improve clinical outcomes and may even
263
lower the quality of life.
264
basis, the GP guidelines advise doing so on a time-contingent basis.
RI PT
23
265
33,34
, since imagery may lead to higher
While analgesics were prescribed to 44.5% of the patients on a pain-contingent 33
TE D
35,36
M AN U
SC
22
Strengths and limitations
267
This is the first study that compares current usual care with the advised stratified-SBT approach in the
268
Netherlands, and provides insight into the question regarding whether specific implementation of stratified
269
care is required or whether usual care already properly categorizes patients.
270
Our results have limited generalizability for the group of high-risk patients because this group was quite small.
271
At the moment, the SBT approach is only available for LBP patients, and so far, no study has been conducted
272
on an SBT approach for patients with NP; however, the conditions are rather comparable. Our categorization
273
of usual care was based on the information that PTs provided. They wrote their (primary) treatment goals and
274
means, and we may have missed certain psychosocial factors that were addressed but not reported in the
275
patient files.
AC C
EP
266
276 277
In the UK, the SBT has the potential to predict persisting disability in clinical practice. We found differences
278
between usual care and the advised stratified care, since most patients received the therapy advised for
ACCEPTED MANUSCRIPT 279
medium-risk patients. Educating clinicians on the SBT approach, especially for the low and high-risk groups, is
280
necessary for this approach to be effective in reducing pain, function and sick leave in the Netherlands.
281
However, prior to making efforts to change the working method, it is essential to determine what causes
282
clinicians to treat patients as they do. A qualitative study should be undertaken to gain insight.
RI PT
283 Conclusion
285
Current Dutch usual care does not correspond to the advised stratified-care approach based on the SBT for
286
patients with non-specific LBP or NP. Clinicians tend to treat the majority of their patients as medium-risk
287
patients even though most of them were found to be low-risk patients.
SC
284
289
Acknowledgments
290
DELETED FOR REVIEWERS
M AN U
288
292
Authors’ contributions
293
DELETED FOR REVIEWERS
294 Conflict of Interest
296
DELETED FOR REVIEWERS
298 299
AC C
297
EP
295
TE D
291
ACCEPTED MANUSCRIPT 300
References
301
1.
Vos T, Flaxman AD, Naghavi M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289
302
diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010.
303
Lancet. 2012;380(9859):2163-2196. doi:10.1016/S0140-6736(12)61729-2. 2.
Côté P, Cassidy JD, Carroll L. The treatment of neck and low back pain: who seeks care? who goes
RI PT
304 305
where? Med Care. 2001;39(9):956-967. http://www.ncbi.nlm.nih.gov/pubmed/11502953. Accessed
306
October 22, 2015.
308 309
seeking. J Manipulative Physiol Ther. 2004;27(5):327-335. doi:10.1016/j.jmpt.2004.04.006. 4.
310 311
Walker BF, Muller R, Grant WD. Low back pain in Australian adults. Health provider utilization and care
SC
3.
Mannion AF, Wieser S, Elfering A. Association Between Beliefs and Care-Seeking Behavior for Low Back
M AN U
307
Pain. Spine (Phila Pa 1976). 2013;38(12):1016-1025. doi:10.1097/BRS.0b013e31828473b5. 5.
Kamper SJ, Maher CG, Hancock MJ, Koes BW, Croft PR, Hay E. Treatment-based subgroups of low back pain: A guide to appraisal of research studies and a summary of current evidence. Best Pract Res Clin
313
Rheumatol. 2010;24(2):181-191. doi:10.1016/j.berh.2009.11.003.
314
6.
TE D
312
van Tulder M, Becker A, Bekkering T, et al. Chapter 3 European guidelines for the management of acute nonspecific low back pain in primary care. Eur Spine J. 2006;15(S2):s169-s191.
316
doi:10.1007/s00586-006-1071-2. 7.
318 319
Bier JD, Scholten-Peeters W, Pool JJ, Staal JB, Tulder MW van, Verhagen AP. KNGF-Richtlijn Nekpijn.; 2016.
8.
320
AC C
317
EP
315
Koes BW, van Tulder M, Lin C-WC, Macedo LG, McAuley J, Maher C. An updated overview of clinical guidelines for the management of non-specific low back pain in primary care. Eur Spine J.
321
2010;19(12):2075-2094. doi:10.1007/s00586-010-1502-y.
322
9.
NHG. NHG-Standaard Aspecifieke Lagerugpijn.; 2005.
323
10.
Staal JB, Hendriks EJM, Heijmans M, et al. KNGF-richtlijn Lage rugpijn. 2013.
324
11.
Hill JC, Dunn KM, Lewis M, et al. A primary care back pain screening tool: Identifying patient subgroups
ACCEPTED MANUSCRIPT 325 326
for initial treatment. Arthritis Rheum. 2008;59(5):632-641. doi:10.1002/art.23563. 12.
327 328
Foster NE, Hill JC, Hay EM. Subgrouping patients with low back pain in primary care: are we getting any better at it? Man Ther. 2011;16(1):3-8. doi:10.1016/j.math.2010.05.013.
13.
Abedi M, Manshadi FD, Khalkhali M, et al. Translation and validation of the Persian version of the STarT Back Screening Tool in patients with nonspecific low back pain. Man Ther. 2015:1-5.
330
doi:10.1016/j.math.2015.04.006. 14.
332 333
Pilz B, Vasconcelos RA, Marcondes FB. The Brazilian version of STarT Back Screening Tool – translation , cross-cultural adaptation and reliability *. 2014;18(5):453-461.
15.
SC
331
RI PT
329
Luan S, Min Y, Li G, et al. Cross-cultural Adaptation, Reliability, and Validity of the Chinese Version of the STarT Back Screening Tool in Patients With Low Back Pain. Spine (Phila Pa 1976). 2014;39(16):E974-
335
9. doi:10.1097/BRS.0000000000000413.
336
16.
M AN U
334
Piironen S, Paananen M, Haapea M, et al. Transcultural adaption and psychometric properties of the STarT Back Screening Tool among Finnish low back pain patients. Eur Spine J. February 2015.
338
doi:10.1007/s00586-015-3804-6.
339
17.
TE D
337
Bruyere O, Demoulin M, Brereton C, et al. Translation validation of a new back pain screening questionnaire (the STarT Back Screening Tool) in French. Arch Public Heal. 2012;70(1):12.
341
doi:10.1186/0778-7367-70-12. 18.
343
Back Screening Tool” (SBST) in different subgroups. Aten primaria / Soc Española Med Fam y
344 345
Comunitaria. 2011;43(7):356-361. doi:10.1016/j.aprim.2010.05.019.
19.
346 347
350
Morso L, Albert H, Kent P, et al. Translation and discriminative validation of the STarT Back Screening Tool into Danish. Eur Spine J. 2011;20(12):2166-2173. doi:10.1007/s00586-011-1911-6.
20.
348 349
Gusi N, del Pozo-Cruz B, Olivares PR, Hernández-Mocholi M, Hill JC. The Spanish version of the “STarT
AC C
342
EP
340
Bier JD, Ostelo RWJG, Hooff ML van, et al. Validity and reproducibility of the STarT Back Tool (Dutch version) in low back pain patients in primary care. PTJ - pending Publ. 2016.
21.
Bier JD, Ostelo RWJG, Koes BW, Verhagen AP. Validity and reproducibility of the modified STarT Back Tool (Dutch version) for patients with neck pain in primary care. Unpubl Obs. 2016.
ACCEPTED MANUSCRIPT 351
22.
Hill JC, Whitehurst DGT, Lewis M, et al. Comparison of stratified primary care management for low
352
back pain with current best practice (STarT Back): a randomised controlled trial. Lancet (London,
353
England). 2011;378(9802):1560-1571. doi:10.1016/S0140-6736(11)60937-9.
354
23.
Murphy SE, Blake C, Power CK, Fullen BM. Comparison of a Stratified Group Intervention (STarT Back) With Usual Group Care in Patients With Low Back Pain: A Nonrandomized Controlled Trial. Spine (Phila
356
Pa 1976). 2016;41(8):645-652. doi:10.1097/BRS.0000000000001305.
357
24.
RI PT
355
Hjermstad MJ, Fayers PM, Haugen DF, et al. Studies comparing Numerical Rating Scales, Verbal Rating Scales, and Visual Analogue Scales for assessment of pain intensity in adults: a systematic literature
359
review. J Pain Symptom Manag. 2011;41(6):1073-1093. doi:10.1016/j.jpainsymman.2010.08.016. 25.
361 362
Vernon H, Mior S. The Neck Disability Index: a study of reliability and validity. J Manip Physiol Ther.
M AN U
360
SC
358
1991;14(7):409-415. http://www.ncbi.nlm.nih.gov/pubmed/1834753. 26.
Brouwer S, Kuijer W, Dijkstra PU, Goeken LN, Groothoff JW, Geertzen JH. Reliability and stability of the Roland Morris Disability Questionnaire: intra class correlation and limits of agreement. Disabil Rehabil.
364
2004;26(3):162-5. doi:10.1080/09638280310001639713.
365
27.
TE D
363
Roland M, Morris R. A study of the natural history of low-back pain. Part II: development of guidelines for trials of treatment in primary care. Spine (Phila Pa 1976). 1983;8(2):145-150. doi:10.1007/978-1-
367
4471-5451-8_59. 28.
369
pain and its relation to behavioral performance. Pain. 1995;62(3):363-372.
370 371
http://www.ncbi.nlm.nih.gov/pubmed/8657437.
29.
372 373
Vlaeyen JW, Kole-Snijders AM, Boeren RG, van Eek H. Fear of movement/(re)injury in chronic low back
AC C
368
EP
366
Sullivan MJL, Bishop SR, Pivik J. The Pain Catastrophizing Scale: Development and validation. Psychol
Assess. 1995;7(4):524-532. doi:Doi 10.1037/1040-3590.7.4.524.
30.
Salen BA, Spangfort E V, Nygren AL, Nordemar R. The Disability Rating Index: an instrument for the
374
assessment of disability in clinical settings. J Clin Epidemiol. 1994;47(12):1423-1435.
375
http://www.ncbi.nlm.nih.gov/pubmed/7730851.
376
31.
Hay EM, Dunn KM, Hill JC, et al. A randomised clinical trial of subgrouping and targeted treatment for
ACCEPTED MANUSCRIPT 377
low back pain compared with best current care. The STarT Back Trial Study Protocol. BMC
378
Musculoskelet Disord. 2008;9:58. doi:10.1186/1471-2474-9-58. 32.
380 381
de Vet HCW, Mokkink LB, Terwee CB, Hoekstra OS, Knol DL. Clinicians are right not to like Cohen’s kappa. Bmj-British Med J. 2013;346(April):f2125. doi:Artn F2125Doi 10.1136/Bmj.F2125.
33.
Chavannes AW, Mens JMA, Koes BW, et al. NHG-Standaard Aspecifieke lagerugpijn | NHG. Huisarts
RI PT
379
382
Wet. 2005;48(3):113-123. https://www.nhg.org/standaarden/volledig/nhg-standaard-aspecifieke-
383
lagerugpijn#note-15. Accessed April 22, 2016. 34.
van Tulder MW, Koes BW. Evidence-Based Handelen Bij Lage Rugpijn.; 2013.
SC
384
http://download.springer.com/static/pdf/687/bfm%25253A978-90-368-0277-
386
2%25252F1.pdf?originUrl=http%253A%252F%252Flink.springer.com%252Fbook%252Fbfm%253A978-
387
90-368-0277-
388
2%252F1&token2=exp=1461316561~acl=%252Fstatic%252Fpdf%252F687%252Fbfm%2525253A978-
389
90-368-0277-2%2525252F1.pdf%253Forigi. Accessed April 22, 2016.
390
35.
M AN U
385
Jarvik JG, Hollingworth W, Martin B, et al. Rapid magnetic resonance imaging vs radiographs for patients with low back pain: a randomized controlled trial. JAMA. 2003;289(21):2810-2818.
392
doi:10.1001/jama.289.21.2810.
393
36.
TE D
391
Karel YHJM, Verkerk K, Endenburg S, Metselaar S, Verhagen AP. Effect of routine diagnostic imaging for patients with musculoskeletal disorders: A meta-analysis. Eur J Intern Med. 2015;26(8):585-595.
395
doi:10.1016/j.ejim.2015.06.018.
397
AC C
396
EP
394
ACCEPTED MANUSCRIPT Figure 1; patient flow
399
Table 1; specific agreement
400
Table 2; baseline characteristics
401
Table 3; specific agreement analysis between baseline risk profile and treatment as provided
402
Table 4a; treatment analysis for low back pain
403
Table 4b; treatment analysis for neck pain
RI PT
398
AC C
EP
TE D
M AN U
SC
404
ACCEPTED MANUSCRIPT Table 1, specific agreement*
Low
Medium
High
Low
(A)
(B)
(C)
Medium
(D)
(E)
(F)
High
(G)
(H)
(I)
RI PT
* “Low-risk” A/(A+(B+C+D+G)/2) “Medium-risk” E/(E+(B+H+D+F)/2)
EP
TE D
M AN U
SC
“High-risk” I/(I+(C+F+G+H)/2)
AC C
Baseline (T0)
Follow-up (T1)
ACCEPTED MANUSCRIPT Table 2, baseline characteristics* Total
GP
Neck (n=100) Back (n=184)
†
PT
Neck (n=26)
Back (n=77)
‡
Neck (n=74) Back (n=107)
Female
65 (65.0)
103 (56.0)
17 (65.4)
42 (54.5)
48 (64.9)
61 (57.0)
Age in years, mean (SD)
45.6 (14.3)
44.7 (14.6)
45.4 (12.5)
40.7 (14.7)
45.7 (14.9)
47.5 (13.9)
Low
58 (58.0)
96 (52.2)
13 (50.0)
42 (54.5)
45 (60.8)
54 (50.5)
Medium
37 (37.0)
70 (38.0)
11 (42.3)
29 (37.7)
26 (35.1)
41 (38.3)
High
5 (5.0)
18 (9.8)
2 (7.7)
6 (7.8)
3 (4.1)
12 (11.2)
<1 month
27 (27.0)
53 (28.8)
7 (26.9)
15 (19.5)
20 (27.0)
38 (35.5)
1 to 3 months
18 (18.0)
26 (14.1)
5 (19.2)
9 (11.7)
13 (17.6)
17 (15.9)
>3 months
55 (55.0)
105 (57.1)
14 (53.8)
53 (68.8)
41 (55.4)
52 (48.6)
||
5.5 (1.9)
5.9 (1.8)
5.4 (1.9)
5.7 (1.9)
§
¶
Disability (NDI) , mean (SD)
14.1 (6.8)
#
9.5 (5.9)
Disability (RDQ) , mean (SD) **
Fear (TSK) , mean (SD)
14.2 (8.8)
32.2 (5.9)
††
Catastrophizing (PCS) , mean (SD) 12.7 (10.0)
6.0 (1.8)
14.1 (6.1)
8.9 (5.7)
10.0 (6.1)
34.8 (7.1)
33.8 (5.6)
35.2 (6.3)
31.7 (5.9)
34.4 (7.6)
13.8 (10.3)
16.5 (12.4)
14.6 (11.2)
11.4 (8.7)
13.1 (9.6)
17 (23.6)
23 (23.2)
EP
TE D
Referral * Values are numbers (percentage) unless otherwise indicated. † General practitioner ‡ Physical therapist § SBT = STarT Back Tool (0-9) || Pain intensity is measured on a Numeric Pain Rating Scale (0-10) ¶ NDI = Neck Disability Index (0-50) # RDQ = Roland Disability Questionnaire (0-24) ** TSK = Tampa Scale of Kinesiophobia (17-63) †† PCS = Pain Catastrophizing Scale (0-65)
AC C
5.6 (1.9)
M AN U
Pain intensity , mean (SD)
SC
Episode duration
RI PT
SBT risk profile
ACCEPTED MANUSCRIPT Table 3, Specific agreement analysis between baseline risk profile and treatment as provided*
Treatment profile Medium
High
13 7 1
81 55 14
1 2 1
22.4% 51.4% 10.0%
2 3 1
55 30 4
0 3 0
6.3% 48.0% N/A
RI PT
Neck pain (n=98) Low-risk Medium-risk High-risk
Specific Agreement
Low
SC
Baseline Low Back pain (n=175) Low-risk Medium-risk High-risk
AC C
EP
TE D
M AN U
* Analysis based on patients’ baseline-data and clinicians treatment-data. Specific agreement is interpreted as the proportion of patients that had that specific risk profile on either baseline or in the treatment profile.
ACCEPTED MANUSCRIPT
Mediumrisk High-risk
Under treated Rightfully treated Over treated
N (%) 9 (12.0) 65 (86.7) 1 (1.3) 6 (11.3) 45 (84.9) 2 (3.8) 0 (0.0) 13 (92.9) 1 (7.1) 19 (13.4) 55 (38.7) 68 (47.9)
SC
Treatment Low Medium High Low Medium High Low Medium High
M AN U
Low Back pain N=142
Baseline Low-risk
Persisting Disability N (%) 2 (22.2) 13 (20.0) 1 (100.0) 3 (50.0) 15 (33.3) 2 (100.0) 0 (N/A) 7 (53.8) 1 (100.0) 10 (52.6) 18 (32.7) 16 (23.5)
RI PT
Table 4a, treatment analysis for low back pain*
* Analysis based on patients’ baseline-data and follow-up data, and clinicians treatment-data. ‘Persisting Disability’ is the amount and percentage of patients with persisting disability in the corresponding treatment group
Neck pain N=92
Baseline Low-risk
Treatment Low Medium High Low Medium High Low Medium High
EP
Mediumrisk
TE D
Table 4b, treatment analysis for low neck pain*
AC C
High-risk
Under treated Rightfully treated Over treated
(%) 2 (3.7) 52 (96.3) 0 (0.0) 3 (8.8) 28 (82.4) 3 (8.8) 1 (25.0) 3 (75.0) 0 (0.0) 7 (7.6) 30 (32.6) 55 (59.8)
Persisting Disability N (%) 0 (0.0) 7 (13.5) 9 (N/A) 0(0.0) 10 (35.7) 3 (100.0) 1 (100.0) 2 (66.7) 0 (N/A) 3 (42.9) 10 (33.3) 19 (34.5)
* Analysis based on patients’ baseline-data and follow-up data, and clinicians treatment-data. ‘Persisting Disability’ is the amount and percentage of patients with persisting disability in the corresponding treatment group
ACCEPTED MANUSCRIPT Figure 1; Patient flow
Back pain N=184
Neck pain N=100
RI PT
Baseline data N=284
PT N=100
Analyzed PT’s N=108
GP N=21
Analyzed GP’s N=26
AC C
EP
TE D
Analyzed GP’s N=75
Both N=8
Both N=5
PT N=72
M AN U
GP N=67
SC
No GP/PT Follow-up N=2
No GP/PT Follow-up N=9
Analyzed PT’s N=77